sorry dell....

[ribitch]

I had to post this. I know I will get flamed and probably beat with a rubber hose, but it needs its due.

linkage (http://www.newsfactor.com/perl/story/18421.html)

[SnowSurfer]

Apple CEO Steve Jobs is well-known for performing "bake-off" speed demonstrations during his Macworld Expo keynote presentations. The tests generally involve Apple's latest Power Mac G4 pitted against a similarly configured Pentium system -_- and the Mac always handily beats its Wintel competitor.

Well of course the apple systems are running slow so if u put a equally configured pc of course the apple is going to win because its top quality is last years pc top quality.

As tested, the Dell 1650 with 108 GB of hard drive space carries a price tag of about US$4900, while the Xserve with 120 GB of hard drive space comes in at about $4500.

ok yeah the xserve has to be priced less to even have a chance to compete. how many people do u see buying them? not many...hmm wierd.

thats my .02 cents

[ribitch]

Originally posted by SnowSurfer
Apple CEO Steve Jobs is well-known for performing "bake-off" speed demonstrations during his Macworld Expo keynote presentations. The tests generally involve Apple's latest Power Mac G4 pitted against a similarly configured Pentium system -_- and the Mac always handily beats its Wintel competitor.

Well of course the apple systems are running slow so if u put a equally configured pc of course the apple is going to win because its top quality is last years pc top quality.

As tested, the Dell 1650 with 108 GB of hard drive space carries a price tag of about US$4900, while the Xserve with 120 GB of hard drive space comes in at about $4500.

ok yeah the xserve has to be priced less to even have a chance to compete. how many people do u see buying them? not many...hmm wierd.

thats my .02 cents

ok, as for comparisons, its not equally rated in terms of MHz, Apple will compares its fastest system to the fastest AVAILABLE intel based systems. So its not last years top quality system, its the currently available top quality. Apple will publish the system specs of what it competes against. Your poiunt here is invalid. sorry.

xserve has only been on the market for a month now(if that). They just started shipping them this week. Over 4000 have been sold and shipped. Not bad in my opinion, since servers arent sold as often as a desktop system. So your point on how many people are using them again is invalid.

The price of the xserve was high in many people opinion, yet when it was compared to an almost equally priced dell, it outperformed (even though the dell cost more). People say apple systems are overpriced, but this proves that the apple has a better price tag than the dell.

[Bires]

I remember when the G3 came out and MacWorld and a few others benched it against the Pentium Coppermine....and the Mac smoked the pentium. It's because they screw the PC with crappy RAM, and run unoptimized drivers and all sorts of unfair testing conditions. When Ziff Davis replicated the tests with fair matched systems...you know the results...

Apple:Market share? We don't need no market share!

[Ladogaboy]

I would trust those tests if they were a little more up-front about the complete hardware configuration. The number of processors and the size the harddrive are just the tip of the iceberg.

The Mac server might be legitimately faster than the Dell server, but until they give the complete break down of the systems, I'll take it with a grain of salt.

I have an update to this, but I'll leave what I was originally going to say because it is still valid.

Well, I finally found a link (http://www.xinet.com/benchmarks/benchmarks.2002/index.html) to the hardware configurations.

1) It still doesn't list what processors the Mac is using.

2) The Dell server is using 2 1.4GHz P3s... definitely state of the art. :rolleyes:

3) The Mac is using 2 Gigabit NICs while the Dell is only using one, and of course, that has nothing to do with thoroughput.

[ribitch]

from xinet

Test with typical prepress applications
In a typical prepress workflow, the heaviest client/server file-transfer traffic is generated by Macintosh operators using image-editing applications such as Adobe Photoshop. Xinet’s Benchmarked Configurations measure network throughput by Macintosh clients opening and saving Photoshop files stored on the server. This yields a very close representation of the transfer rates that users will see in real-world conditions.


Test with concurrent multiple clients
Real-world servers typically have more than one client, and measuring a server’s performance with a solitary client doesn’t predict how well that server will perform under multiple-client load. For these benchmark tests, Xinet chose to overload the server’s capacity on a single network adapter and single RAID system. Xinet estimates that under normal conditions Photoshop operators will perform a Save or Open on high-resolution files for no more than one-tenth of their total work time. Therefore, ten concurrent Macintosh test clients, continuously calling for file “reads” and “writes” from the server, represent one hundred real-world clients, under average file-server use (as shown by the “Simulated Load” column in the Open/Save tests).

Use large files
Prepress client/server access is typically dominated by sequential “reads” and “writes” of large (100+ MB) files. Using small files for benchmarking purposes tests little more than the caching ability of the client or the server. All network benchmarks were conducted using a 100MB file on each client.

----------------------------------------------------------

As you can see, these are real world server tests. Not photoshop tests where one test is better on CISC and the other is better on RISC.

Since these are business servers, it makes sense to test them for things of this nature, not photoshop rendering. These were completed by an independent form, so the RAM wasn't switched, and the drivers werent buggy drivers.

Sure, the suns beat apple like you wouldnt believe, but there is also a steep price on those, and they also have way more processors.

[ribitch]

Originally posted by Ladogaboy
I would trust those tests if they were a little more up-front about the complete hardware configuration. The number of processors and the size the harddrive are just the tip of the iceberg.

The Mac server might be legitimately faster than the Dell server, but until they give the complete break down of the systems, I'll take it with a grain of salt.

I have an update to this, but I'll leave what I was originally going to say because it is still valid.

Well, I finally found a link (http://www.xinet.com/benchmarks/benchmarks.2002/index.html) to the hardware configurations.

1) It still doesn't list what processors the Mac is using.

2) The Dell server is using 2 1.4GHz P3s... definitely state of the art. :rolleyes:

3) The Mac is using 2 Gigabit NICs while the Dell is only using one, and of course, that has nothing to do with thoroughput.

1) xserve is dual gigahertz G4's.
2) I am assuming they used PII's to keep the price similar to the mac, since it wasnt the high end Xserve. Both dell and apple servers used were their midrange servers.

3)

All of the platforms Xinet tested used the same network topology: a single Copper Gigabit network with a single switch. Also, all tests were performed on a single RAID/disk on a single controller. This is a significant factor, especially when measuring writing performance. This round of benchmarks proved that Copper Gigabit technology is reliable

[xsiled2]

ribitch you sir are pathetic.

a reatrded gerbal with one arm could make a better pc with $5k then that crappy out dated dell. where did they find this price..

whats wrong, to chicken to take it up with 2 xeons, EVERYONE knows pIII's arnt server procs.

:bigmouth: :bigmouth: :bigmouth:

[ribitch]

Originally posted by xsiled2
ribitch you sir are pathetic.

a reatrded gerbal with one arm could make a better pc with $5k then that crappy out dated dell. where did they find this price..

whats wrong, to chicken to take it up with 2 xeons, EVERYONE knows pIII's arnt server procs.

:bigmouth: :bigmouth: :bigmouth:

i knew i would here from you...

the dell and the apple as previously stated are midrange systems. Sure you could build a dual processor Xeon system, and get the price well over 5k.

A xeon also makes the server a 2U tall server, unlike apples 1U. If you wanted to go dual xeon, the price jumps to 7500 for just the basic system with 1 GB RAM. Plus you would need to tack on more similar harddrives to the rest of the ststems. I didnt design these tests.

The point at posting this article was to show that a SIMILARLY priced apple outperforms a dell in a real world situation. It wasnt to show that apple can beat a midrange dell with their midrange server. Yes, both companies have better faster systemns than the ones tested, but most companies dont always order the most expensive fastest server. They order what performs best for their money, or something within budget. A 10k dollar dual xeon isnt in most companies budget, but a system of half that price is.

You could ideally order 2 of the apples as tested and fit them in the same space as the xeon would take up for teh same price. The apple systems would be able to run cirlces around the one dell.

So bitch all you want about how the dell wasnt a dual xeon. I dont care.

[xsiled2]

actually if you go with 2 xeon 1.8ghz i comes out to $5100 on 2 hundred bucks more with a slight loss of hard drive. comparing 2 older cpu's against the mothership of your crappy smear based company is less then poor. im surprised they even tested pIII's againsts it, im sure they could have saved money and boughten a dual celeron setup so they could say the mac was better. HOW LAME.


Originally posted by ribitch

So bitch all you want about how the dell wasnt a dual xeon. I dont care.

SO WHATS THE POINT OF THE COMPARISON? your new pos versus 2 cpus made about a year ago.

[ribitch]

Originally posted by xsiled2
actually if you go with 2 xeon 1.8ghz i comes out to $5100 on 2 hundred bucks more with a slight loss of hard drive. comparing 2 older cpu's against the mothership of your crappy smear based company is less then poor. im surprised they even tested pIII's againsts it, im sure they could have saved money and boughten a dual celeron setup so they could say the mac was better. HOW LAME.



SO WHATS THE POINT OF THE COMPARISON? your new pos versus 2 cpus made about a year ago.

1st, xinet isnt in any way related to apple. get that straight. This isnt apple toughting themselves as better. Its simply an INDEPENDENT company testing servers.

I went with the fastest xeon to get you to quit bitching. The servers were set up with certain minimum specs. So again, there is a price increase.

The point of the comparisons was to show how systems performed using xinets prepress software across all platforms and what would be a best value for your solution. performance per dollar was in apples corner, sorry.

Let me guess, you are going to start to say "well, who runs photoshop on a server anyways?" Let me clear that up right now. Photoshop is run off of a client, the server is acting as a fileserver. Photoshop isnt being run on the server. You could replace photoshop with any sort of program. They used photoshop because these are tests used by a company that creates prepress server software.

so once again, quit your bitching. get your information correct before you start running your mouth off. Read the system requirements before you say "well, theres a decrease in harddrive, and this, and that..." Or "For x amount more, you cold get this" Remember, the apple is already 400 dollars cheaper than the dell in the first place. You could get a whole lot more system for a few hundered dollars more to match dells price.

[SnowSurfer]

ribitch so u are telling me a new apple imac (or g4) can take on a pentium 4 at 2.53 ghz? i doubt it. ribitch are do u own a apple? are u a apple lover. i used to have an apple and i couldnt play games with it couldnt do anything i jumped over to pc and you know what its great no problems no worries and i can overclock and add stuff. Winxp is stable and there is more win programs too. how many apple overclocking sites is there? apple is trying to make a comeback and failing. apple is like a mouse and dell (and such) are lions. who is going to win the mouse or the lion?

[Cantacuzene]

Ribitch, you make some valid points, but so does Xsiled. You seem to be so in love with Apple that no matter what is said you never regard it and rationalize it some other way. Maybe you should perhaps accept that APPLES MAY NOT BE THE BEST AT ALL THINGS ALL THE TIME AT ALL PRICES, which is what you seem to subscribe to. Chill out and stop being defensive about it.

For someone who doesnt work for Apple you are suprisingly eager to defend a company that doesnt give a damn about you. If you were under as much attack as Apple is, would they defend you? I think not.

Listening to these endless Mac vs PC arguements is only slightly more interesting and mature than listening to Xbox vs PS2 arguements in the hardforums.

[Bires]

Originally posted by xsiled2
ribitch you sir are pathetic.


Originally posted by SnowSurfer
are u a apple lover.

/me throws hands up...

:cry2: "You're making the babies cry..." :cry2:

[ribitch]

Originally posted by SnowSurfer
ribitch so u are telling me a new apple imac (or g4) can take on a pentium 4 at 2.53 ghz? i doubt it. ribitch are do u own a apple? are u a apple lover. i used to have an apple and i couldnt play games with it couldnt do anything i jumped over to pc and you know what its great no problems no worries and i can overclock and add stuff. Winxp is stable and there is more win programs too. how many apple overclocking sites is there? apple is trying to make a comeback and failing. apple is like a mouse and dell (and such) are lions. who is going to win the mouse or the lion?

i have owned bot and i use both. I am not much of a gamer. I prefer to spend 300 bucks for a console to play games. That console only needs to be upgraded every four to five years to play the latest and greatest games, unlike a pc who needs it every year if not sooner.

The new imac is a lowend system. Plain and simple. Yeah, you can probably get a 2.53 GHz Dell for as much as a loaded iMac after rebates, but it will need to be updates sooner than th imac. Apples maay cost more up front, but the actual cost of ownership is lower (several studies have found this), thge require less maintenance than a windows machine, they last longer before needing upgrades, and they sell for far more than a PC after 2 years of use.

So for a cost perspective, apple is the better choice. Performance perspective, a dual proc PowerMac is my choice for an everyday system.

Sure, there are lots more programs written specifically for PC. These are often system utilities, or games, or anything else programmed in visual studio. However, theres a reason why so many compaines of business software have commited themselves to releaseing an OS X version of their software (the likes of Oracle, M$, Maya, Macromedia, iD, IBM, and symantec). That reason is OS X is the easiest to use Unix OS on the market, apple is the worlds largest Unix vendor, and Apple is turning heads. So your company had a program designed specifically for themselves in visual studio. You cant run it natively, you need to stay windows based or use an emulator. That is your companies fault for not requiring it to be programmed in Java or C/C++. You cant easily port a propietary coded program, but you can port a universal language. Hell, java run son virtually everything. Thats why apple uses Java for WebObjects. You have the simplicity of an IDE like visual studio, but you are not limited to one architecture.

People complain about apple stating they are only good for graphics. These people are mainly PC users. Apple users know that games look great on their systems, they also know that they can do everything a PC user can (with the aid of Virtual PC) and more!

I did not post this as an attempt to convert people. This thread is about fileserver performance. Apple edged out dell in the same price ranged. Plain and simple.

[xsiled2]

Originally posted by ribitch

This thread is about fileserver performance. Apple edged out dell in the same price ranged. Plain and simple.

no this thread is about macs flagship against cheese ball dell's 6 months ago pIII dual system. which is not in anyway the flagship.

[ribitch]

Originally posted by Cantacuzene
Ribitch, you make some valid points, but so does Xsiled. You seem to be so in love with Apple that no matter what is said you never regard it and rationalize it some other way. Maybe you should perhaps accept that APPLES MAY NOT BE THE BEST AT ALL THINGS ALL THE TIME AT ALL PRICES, which is what you seem to subscribe to. Chill out and stop being defensive about it.

For someone who doesnt work for Apple you are suprisingly eager to defend a company that doesnt give a damn about you. If you were under as much attack as Apple is, would they defend you? I think not.

Listening to these endless Mac vs PC arguements is only slightly more interesting and mature than listening to Xbox vs PS2 arguements in the hardforums.

i actually worked with apple for 1.5 years. I plan on getting a job with them once again once I graduatre. I was a narrowminded PC user before I got the apple job. I took it for the free use of a laptop and some extra cash. I ended up finding out that my apple was a better system than my PC. I converted. I still use PCs daily (right now I am using a dell optiplex GX240 at work), but i prefer the apple. I know they are not the best solution for everybody, but I often hear people that dont know what they ar etalking about, and therefore I will voice my opinion.

I have the viewpoint of cost effectiveness of the hardware with the longetivity of the hardware. I dumped at least 600.00 anually into my system trying to stay within a semi recent hardware configuration.

Personally, i love bickering back and forth like this. It actually gets interesting hearing people skim an article or something, and start voiceing their oopinions before they read the entire article or whatever. I especially like bickering with people who have never tried uing another type of system and just slam it like you wouldnt believe (like myself 2 years ago). I was many of your positions before. I know how good it feels to slam apple. Its almost like messing with a n00b. But messing with PC users is even more enjoyable, because I am aware of all the things they dont know, and I can correct them and things of that nature. Its fun to make somebody who is clueless about something even though they think they arent look liek a complete ass.

[xsiled2]

so now your a narrow minded mac user.

[ribitch]

Originally posted by xsiled2
so now your a narrow minded mac user.

not quite. sorry.

[jase71]

Originally posted by xsiled2


no this thread is about macs flagship against cheese ball dell's 6 months ago pIII dual system. which is not in anyway the flagship.

If this is true, and it was Apple's flagship against an archaic Dell... and yet the Apple system was cheaper...

Then isn't Dell the one with the outrageous prices, and not Apple? :P

[xsiled2]

no jase, thats the 6 months ago price.

the dually xeon is only $200 more now then what that WAS.

dell still sucks.

[ribitch]

Originally posted by jase71


If this is true, and it was Apple's flagship against an archaic Dell... and yet the Apple system was cheaper...

Then isn't Dell the one with the outrageous prices, and not Apple? :P

thanks. I needed some help in here. I never even thought of that.

So xsiled2, why is the old dell system more expensive? It is 6 months old remember? And it is using last years processor.

[ribitch]

Originally posted by xsiled2
no jase, thats the 6 months ago price.

the dually xeon is only $200 more now then what that WAS.

dell still sucks.

200 with a cut in harddrive, and you never did say that amount of RAM. we are talking the same specs as the systems tested. 1 GB ram 108 GB HD with RAID and gigabit ethernet. These tests were not 6 months ago. They were released last week i believe. So once again, please try again

[xsiled2]

Originally posted by ribitch

So xsiled2, why is the old dell system more expensive? It is 6 months old remember? And it is using last years processor.

i said 6 months agos price.

dell has unrealistic prices now so the comparison is still crap.

[ribitch]

we are talking about servers here, not home pc's. Most vendors are expensive due to their onsite warranty programs and support services, plus the hardware needs to be extremely reliable. If you get a server, you dont go to your corner PC store and buy parts. You buy from dell, ibm, hp/compaq, sun, and now apple.

that price isnt 6 months ago either. dell may be more expensive, but you are paying for their warranty services and support.

[jase71]

Originally posted by xsiled2


i said 6 months agos price.

dell has unrealistic prices now so the comparison is still crap.

Considering you can't buy anything using the prices from 6 months ago, you have to use Dell's current prices, it doesn't really matter what they cost 6 months ago.

And you can't really blame Apple for Dell's crappy pricing tier.
It's not Apple's fault Dell has price problems.

So it would seem the comparison is valid.

I just priced out a Dell PowerEdge 2650 on Dell's site. A dual 2.2 gig P4 system is $6,547.00 matching the Apple specs as closely as I can. A dual 2.4 gig machine is $7,147.00 Dropping to a single CPU P4 saves you about $700, but even the single CPU system would cost more than the Apple.

You have to drop to PIII servers to get price competitive with Apple.

I'm not an Apple fan, at least not a fan of their hardware. But that's a heck of a deal on a server, any way you cut it.

[xsiled2]

FINE.

heres a penguin computing server

System Price $ 4816.00
dual 1.8 xeon
1gig 2100 DDR (macs use sd)
3 36.4 15,000 scsi

2 years warranty
3 years on parts.

up ur nose with a rubber hose.

[xsiled2]

OH and that ones U1 so :P

[jase71]

Good deal on a server, xsiled... although that's for a
Redhat Linux server, not a Windows server. I'd certainly think a Linux server would be an excellent way to go. ;) You'd have to add
a nice expensive Windows license to the price for a MS box, though.

Now find ribitch some benchies on it. :P

[xsiled2]

why put windows on it, the mac doesnt have an added price due to its using of its own OS.

linux servers and win servers dont differ much just command and compatibility differences that are worked out by their own applications.

[ribitch]

Originally posted by xsiled2
FINE.

heres a penguin computing server

System Price $ 4816.00
dual 1.8 xeon
1gig 2100 DDR (macs use sd)
3 36.4 15,000 scsi

2 years warranty
3 years on parts.

up ur nose with a rubber hose.

Macs actually use DDR in their servers :stupid:

whatabout gigabit? firewire? Hardware monitoring? Operating System? Client limit(OS wise, not hardware limitation)? Included Software?

still more than the mac, and you dont get dells reputation. How about service for parts? same day air? next day? Who delivers and installs?

for the xserve:
10/100/1000 (option gigabit fibre)
2 firewire (fast easy datatransfers without giving network bandwidth)
hardware monitoring with indicators on case and software
OS X unlimited client
WebObjects, Quicktime Server, NetBoot
4 hour onsite response during business hours otherwise next day

Buying a server isnt like getting a workstation or dekstop system. If you have a failure, you need immediate replacement.

now how do they compare?

[Jeffbx]

That was a very interesting article...

It's kind of neat to see Apple trying the waters in the server market. They've been missing that for years. Their pricepoint is also very competitive. In the server market, Dell is among the cheapest you can buy. Try configuring a similarly equipped Compaq or IBM - they'll easily be a grand higher.

Now, can Apple compete directly against Dell for market share? Probably not. I see the MacOS as really a niche market. It'll probably be nice for the print layout shops and the web designers who still use Mac, but I don't see it as a mainstream server for general business apps. This is because that market is already dominated by Wintel - UNIX and even Linux are still niche players in this area. One of tham main barriers of entry are the admins... you buy platforms that your admins can implement efficiently, you don't buy a new platform based on hardware cost.

I'd like to get my hands on one to play with, tho... see how it holds up against mainstream UNIX platforms like IBM, Sun, HP and SGI.

[jase71]

Originally posted by xsiled2
why put windows on it, the mac doesnt have an added price due to its using of its own OS.


I only mentioned it because the original Dell machine had Win2K on it... that's what it was benched with.

OS plays a major role in performance... even the original Dell might have fared much better in the benchies if it had had Linux rather than Win2K. Linux does some thing much better than Win2K (and some things worse, as well).

But I'm all for running Linux, and leaving Win2K on the shelf!

[xsiled2]

FORM FACTOR
1U Rackmount Form Factor (1.75")
PROCESSOR(S)
Processor Number 2
Processor Type Dual Intel® Xeon Processors
Processor Speed Options 1.8 GHz, 2.0 GHz, 2.2 GHz
Processor Cache Size 512KB Advanced Transfer L2 Cache
MOTHERBOARD
Chipset Intel E7500
Front Side Bus Speed 400 MHz
MEMORY
Memory Type PC2100 ECC Registered DDR SDRAM - Interleaved
Memory Capacity Six DIMM Slots for up to 6 GB
INTERNAL STORAGE CONTROLLER
SCSI Controller Options Two Ultra-160 SCSI Channels
IDE Controller Options Two UDMA 100/66/33 bus master/EIDE Channels
STORAGE SYSTEM
CD-Rom Devices 24x IDE Slim CD-Rom Drive
Hard Drive Type SCSI
Hard Drive Bays
3.5 inch: 3 Hot Swap low profile (1")
Hard Drive Speeds 10,000 RPM / 15,000 RPM
Hard Drive Sizes 18.3, 36.4, 73 GB (10k RPM) and 18.3, 36.4 GB (15k RPM)
Hard Drive Options Single or Dual Channel Ultra-160 RAID Cards
External Drive Enclosure Optional 10 SCSI LVD Hard Drive Enclosures (Single and Dual Backplane)
NETWORKING
Ethernet Controller Onboard Intel® 82544GC 10/100/1000 and 82550 10/100 controllers
On-Board LAN One 10/100/1000 RJ-45 Interface and One 10/100 RJ-45 Interface
Networking PCI Card Options
10/100 Mbps: Single, Dual and Quad RJ-45 Port Cards
1000 Mbps: Single Fiber (SX) Port or Copper (T) Port Cards
PCI EXPANSION SLOTS
Number of Slots Single Full - Length PCI Slot
Slot Speed 64-bit / 66 MHz via riser card
EXTERNAL I/O INTERFACES
Serial Ports 1
PS/2 Keyboard and Mouse Ports Yes
USB Ports 2
GRAPHICS SUBSYSTEMS
Graphics System Integrated ATI Rage XL AGP VGA
Memory 4 MB Frame Buffer
SUPPORTED OPERATING SYSTEM
Red Hat 7.2 and 7.3
POWER SYSTEM
Power Supply Size 460 Watt
DC Output +5V = 22A +12V= 8A +3.3V = .5A -5V = 0.2A -12V = 14A +5VSB = 1A
COOLING
Internal System Fan(s) Six 1.75" x 1.75" fans for front-to-back pressure-based air flow cooling
Power Supply Fan (s) Two 1.75" x 1.75" fans
MOUNTING HARDWARE
Rackmount Rails Ball Bearing Slide Rails
Brackets Front and Rear Mounting Brackets
SYSTEM DIMENSIONS & WEIGHT
Height 1.75"
Width 16.8"
Depth 24.5"
WARRANTY
2 Year Standard

-------------------------------

the mac is over 7k on their site...

[ribitch]

Originally posted by Jeffbx
That was a very interesting article...

It's kind of neat to see Apple trying the waters in the server market. They've been missing that for years. Their pricepoint is also very competitive. In the server market, Dell is among the cheapest you can buy. Try configuring a similarly equipped Compaq or IBM - they'll easily be a grand higher.

Now, can Apple compete directly against Dell for market share? Probably not. I see the MacOS as really a niche market. It'll probably be nice for the print layout shops and the web designers who still use Mac, but I don't see it as a mainstream server for general business apps. This is because that market is already dominated by Wintel - UNIX and even Linux are still niche players in this area. One of tham main barriers of entry are the admins... you buy platforms that your admins can implement efficiently, you don't buy a new platform based on hardware cost.

I'd like to get my hands on one to play with, tho... see how it holds up against mainstream UNIX platforms like IBM, Sun, HP and SGI.

some good points.

Xserve would fit in well as a fileserver or webserver. It dopesnt need to be a windows or unix authenication server.

webserver
Apache has been benchmarked on OS X and on the Xserve and is considered on of the fastest implementations of it.

Fileserver
can do active directory as well as NFS shares. All this can be done with a user friendly management program. Much easier than that of Samba

The Xserve will be a hit in education, since the management is designed for simplicity. Apple makes its software as user friendly as possibly, and the xserve is no exception. by clicking a few buttons, one can set up a netboot server, print server, file server, webserver, and much more. Many IT people in K-12 were given that job because they knew alittle more than the next person. ince most of these people are often teachers as well, they cannot be babysitting a server, or reading books and books on how to run the system let alone administer it.

Apple is also marketing the xserve as a high end workstation for video professional. I wont get into that though.

[Jeffbx]

Hmmm... only IDE drives, tho? No RAID, either? I don't know about that. They'd be better off with a 2U unit that has some redundancy built in.

[ribitch]

Originally posted by xsiled2


the mac is over 7k on their site... [/B]

that model of the xserve is the highend. we are talking about the midrange with these specs:

Summary
• Dual 1GHz PowerPC G4
• 1GB DDR SDRAM - 2 DIMMs
• 120GB Ultra ATA - 7200rpm - Bay 1
• CD-ROM drive
• Gigabit Ethernet Card
• ATI Graphics Card
• Mac OS X Server, Unlimited License

Subtotal $4,549.00

$7800 will get you the top of the line system loaded with 480 GB of drives and 2 GB of DDR. That ISNT the system that was benchmarked.

[ribitch]

Originally posted by Jeffbx
Hmmm... only IDE drives, tho? No RAID, either? I don't know about that. They'd be better off with a 2U unit that has some redundancy built in.


if talking about xserve SCSI is available, and it is raid capable. I havent seen it anywhere in the details of the server, but it is in the knowledge base

[xsiled2]

Originally posted by ribitch



if talking about xserve SCSI is available, and it is raid capable. I havent seen it anywhere in the details of the server, but it is in the knowledge base

well where is it?

the system i laid out from penguincomputing BLOWS away that mac on hard drive speed due to it being just as much space laid out on 3 indepenant 15k scsi drives which are MUCH MUCH faster then that 7200.....

a mac with the same hard drive setup would be in the 7 thousands making that linux solution not only costs effective but MORE powerful.

[SnowSurfer]

ribitch: i cant seem to find the server used in the test on apples site can u please give me a link to it. or are they not selling it and just using it for testing. and ribitch it is about productivity and on this one the gains do not equal or outweigh the risks of putting in a new (mac) server for windows machines

[ribitch]

Originally posted by SnowSurfer
ribitch: i cant seem to find the server used in the test on apples site can u please give me a link to it. or are they not selling it and just using it for testing. and ribitch it is about productivity and on this one the gains do not equal or outweigh the risks of putting in a new (mac) server for windows machines

http://store.apple.com/

click the xserve

select the middle model
change drive 1 to 120 GB
change ram to 1 GB

update price
thats the system

i never stated replacing windows machines with the xserve.

[Speedfreak]

Apple can have this win (whether it is actual or not). They need a trophy in their empty case.

[ribitch]

Originally posted by Speedfreak
Apple can have this win (whether it is actual or not). They need a trophy in their empty case.

i dont think its so empty...
theres an emmy (http://www.apple.com/firewire/)

5 recent IDEA awards (http://www.businessweek.com/magazine/content/02_27/b3790101.htm) (Industrial Design Excellence)

Powerschool has a 2002 Codie (http://www.siia.net/codies2002/winners.htm)

If you want to get into software feats, apple dominates withGenetech Blast. Theres also the Seti@home factor, the iPod, and powerbook.

seems to me apple has several recent awards and envies. sorry

[LegendKiller]

Ok, I would like to point out a few OBJECTIVE observations.

Apple has always "loaded the deck" on their comparisons, and this seems to be no exceptions.

Now, considering the Photoshop "fileserve" you need to find out a few things.

1. Are both systems using 64bit/66mhz PCI slots for their gigabit cards. If one is and the other isn't, the test is invalid since it would give one a throughput advantage. Furthermore, if one doesn't have that subsystem, then it would get bogged down at the 133mb/s limit of PCI33

2. Are both systems using the same HDD's? File serving is EXTREMELY HDD intensive. Those numbers are very fishy to me. Why? Because it says the mac was able to toss out files at 80mb/s. Unless they are using some 20k rpm ATA drive, thats impossible. The highest sustained transfer rate for the fastest ATA drive is somewhere around 50mb/s. Only Cheetah X15's get up to 80mb/s. Again, if you are going to get that high, you have to be operating a higher bus-speed.

3. Ram-the P3 system was probably using SDram, or very old RDRAM which can't even compare to new DDR with a 133mhz bus speed. Again, throughput goes to the Mac

4. The Gigabit card HAS to be the same in order for the test to be valid.


These tests usually lack any scientific validity because these companies do not release full test configuration information, and thus they cannot be duplicated.


There was one interesting project a while ago (when the dual macs came out) from a guy who used his own home cooked version of linux, and a DNA processing program. He TOTALLY leveled the playing field. Even going as far as getting the EXACT same ram, hard drives, NICS, SCSI, and video cards. The only two variables left were the CPU and the mobo.

Guess what? Mac AND intel was trounced by AMD. I really wish I could find the article because it would show all of the mac zealots that what is published isn't always true.

Furthermore, comparing a Mac which sells at 4.5k to a dell that sells at 4.8k isn't exactly fair. Especially considering the economics of the Mac entry into the server market. Mac knows that they have to go for penetration right now, and thus they are pricing their servers towards that. I would bet they are taking a 1k loss on EACH server they are releasing just to gain market share. Once that happens you will see the prices creep up to meet the always-high desktop prices.

LK

[LegendKiller]

Upon further investigation it looks like the dell DOES have SDram, which invalidates the whole test

Furthermore, the pricepoint they have is rediculous, especially considering the Dell has an integrated high-end SCSI controller, which adds more to the cost.

Then, they say that the Dell has a max storage of 108gigs, which is odd considering you can have 120gig drives in it.

Add that to the fact that the MAc has 4 INDEPENDANT (read, master only which reduces I/O interference) IDE channels, which are probably hooked into the 64/66 buses while the Dell is the plain old integrated 32/33 bus.

Then, the drives that they used are almost assuredly NOT the same.

You are comparing an apple to an orange in this case. If any statistician or scientist took a look at these tests he would laugh his ass off and then go buy a dell. Why? Because in doing this Apple shows that its too afraid to compete on a level playing field. I have been through 4 stats classes and I have a BS in psychology from the #1 psychology school in the nation. I have been drilled over and over to eliminate extraneous variables if you want to have a valid test. None of these places do this, and thus even saying that one or the other is better based off of invalid benchmark is not thinking the whole way.

Perhaps I can talk to Darth and we can try and get a TRUE comparison, apple might or might not go for it, and we might have to cook up our own PC. It would be quite interesting to do this type of review.


LK


LK

[ribitch]

off the server topic, but it talks about plug and play systems.

http://www.eweek.com/article2/0,3959,324759,00.asp

it doesnt test XP, but it talks about 1394 plug and play in 98SE, 2k, and OS X. I know xsiled will be like "well XP is much better" and a few others, I know its better, so you dont need to post that.

Does anyone have any articles from a valid source that talk about XP and plug and play? Just curious.

[xsiled2]

i dont trust XP i run 2k and 98se get you head out of your ass you arogant bastard.

[jase71]

Originally posted by LegendKiller


You are comparing an apple to an orange in this case.

So which is more important? Keeping the hardware specs similar? Or at least as similar as possible?

Or keeping the price as close as possible?

We've already established that a similarly configured Dell is much more expensive than the Apple server. The Dell used was chosen because it fell into the same price bracket, not because it matched identically in hardware.

Was the test done to determine how the fastest Apple would compare with the fastest Dell?

Or was the test done to determine how fast a server you could get for ~$4500?

Which one of those two options was the focus might explain the way the comparison was run....

[LegendKiller]

Again, a scientifically invalid test.

Comparing Win2k's driver set (mind you it is geared towards business users and its driver set was created in 1999, while OSX was released AFTER that and is geared as a consumer OS, not to mention the fact that apple designed IEEE 1394, so it should be more up to spec.

Furthermore, considering the older driver set of 2k, and the fact that the company probably wrote crappy drivers, you get problems connecting. THis is compounded by 98SE.

I have always said that people make MS work 10x as hard as ANY other os company. Just think of the multitude of systems it has to work with. With mac there is a finite number of peripherals it connects to, and a finite amount of mobo's and such is uses. However, on the PC you can install windows on so many different mobo's and such that its impossible for MS to keep track of it all. Not to mention the fact that many incompatibilities are a result of inefficient drivers or non-adherent spec'd hardware.

XP-PnP is very good, it has detected every device I have plugged in flawlessly.

Why no responses to my posts?

LK

[LegendKiller]

The pricepoint of the dell is rediculous. You are not even counting the fact it has SCSI integrated and thus it is on a whole different plane of existance. I mean, common. If they actually used SCSI (which I am sure they did not), then the dell would cream the Mac becuase the macs have IDE HDD's. Thus, even with the same price-point you are dummying down the dell and not comparing to equally again.

Its like buying a 2-year old Porche Boxster and then buying a NEW Toyota Camry (And dont forget that you get the Camry at a VERY deep discount because it was JUST released and Toyota is trying to gain market share, so the actual manu price is about 5k HIGHER than what they are selling for), taking the Porche engine out and putting a Geo Metro 3-cyl in and then benchmarking the two. Sure, the Porche is more expensive, then strip it down to the metro engine and its only 25k compared to 19k. Then you compare the price points and guess what? The Camry will win because you bought the porche with a good engine, took it out replaced it with a crappy engine, and then claimed it LOSES because of the crappy engine AND its higher price.

That makes no sense at all.



Originally posted by jase71


So which is more important? Keeping the hardware specs similar? Or at least as similar as possible?

Or keeping the price as close as possible?

We've already established that a similarly configured Dell is much more expensive than the Apple server. The Dell used was chosen because it fell into the same price bracket, not because it matched identically in hardware.

Was the test done to determine how the fastest Apple would compare with the fastest Dell?

Or was the test done to determine how fast a server you could get for ~$4500?

Which one of those two options was the focus might explain the way the comparison was run....

[ribitch]

Originally posted by xsiled2
i dont trust XP i run 2k and 98se get you head out of your ass you arogant bastard.

ok, i didnt know. Once again calm down. I posted that because you always seem to be the most willing to tear apart anything i say.

as for specs, I think the same price range was being achieved, since if you are planning on spending more, you probably will go SGI or SUN since they can provide much more processing power. That may change when 64 bit computing becomes more widely available.

legendkiller
true, 2k is older, but 1394 plug and play works flawlessly in OS 9, even os 8.5.1 when it comes to firewire harddrives. Based on the way the author talked about macs, he probably doesnt use OS 9, therefore didnt comment. Both thoses OS's are older.

sure, the tests are not scientific, but they are realy world tests. Sure, you can script a photoshop test or a quake 3 demo but thats all scripted. Besides, 30 fps in quake is all that is needed. Thats all the eye can see. if its a constant 32 or 99 fps in quake3, it rwally doesnt matter. Graphics complexity does matter though.

correct me if i'm wrong, but MS doesnt write all teh drivers for all the hghardware that windows supports. They include most drivers from other vendors. as far as your apples to oranges, all mobos used are different from sun to apple to dell. yes, each ide channel on the xserve is independent, but nobody told dell they couldnt do that did they? Innovation is what allowed apple to do this.

Scientific tests are nice, but REAL WORLD tests are what matter. The apple and dell appear to be in the same price range.

[jase71]

Originally posted by LegendKiller
The pricepoint of the dell is rediculous. You are not even counting the fact it has SCSI integrated and thus it is on a whole different plane of existance. I mean, common. If they actually used SCSI (which I am sure they did not), then the dell would cream the Mac becuase the macs have IDE HDD's. Thus, even with the same price-point you are dummying down the dell and not comparing to equally again.

First of all, Dell's pricing structure is hardly Apple's problem. If Dell can't equip a comparable server for the same dollar amount, why should they be given the advantage of using a more expensive box for the test?

Secondly, before you rag on the scientific value of the comparison, you might want to read the details.

The Dell server DID use SCSI. It was a Dell 1650 Poweredge server, equipped with 3x36MB drives equipped in a SCSI RAID array.

To answer your other concern, the DELL also has two 64bit/66mhz PCI slots. So you can cross that difference off your list as well.

Details on the Dell 1650 Poweredge are available at:

http://www.dell.com/us/en/bsd/products/model_pedge_2_pedge_1650.htm

[DoPeY5007]

two servers, about the same price, who performed better?

[LegendKiller]

Sure, MS doesn't write its down drivers, It depends on the IHM's to make them, and many times those co's write crappy drivers.

Sure, the price is similar now, but thats only because apple is going for market penetration, not income maximization

Sure, Apple is allowed independant channels, but they are NOT real world examples. They are unfair benchmarks created to spread FUD about PC's. A real world benchmark is comparing 2 like systems at a price point. Not buying 2 same-priced, crippling one in order to make the other win. In a standards-based market like computers you should compare like to like, but muddying this line is where Apple gets its profits.

LK



Originally posted by ribitch


ok, i didnt know. Once again calm down. I posted that because you always seem to be the most willing to tear apart anything i say.

as for specs, I think the same price range was being achieved, since if you are planning on spending more, you probably will go SGI or SUN since they can provide much more processing power. That may change when 64 bit computing becomes more widely available.

legendkiller
true, 2k is older, but 1394 plug and play works flawlessly in OS 9, even os 8.5.1 when it comes to firewire harddrives. Based on the way the author talked about macs, he probably doesnt use OS 9, therefore didnt comment. Both thoses OS's are older.

sure, the tests are not scientific, but they are realy world tests. Sure, you can script a photoshop test or a quake 3 demo but thats all scripted. Besides, 30 fps in quake is all that is needed. Thats all the eye can see. if its a constant 32 or 99 fps in quake3, it rwally doesnt matter. Graphics complexity does matter though.

correct me if i'm wrong, but MS doesnt write all teh drivers for all the hghardware that windows supports. They include most drivers from other vendors. as far as your apples to oranges, all mobos used are different from sun to apple to dell. yes, each ide channel on the xserve is independent, but nobody told dell they couldnt do that did they? Innovation is what allowed apple to do this.

Scientific tests are nice, but REAL WORLD tests are what matter. The apple and dell appear to be in the same price range.

[ribitch]

Originally posted by jase71


First of all, Dell's pricing structure is hardly Apple's problem. If Dell can't equip a comparable server for the same dollar amount, why should they be given the advantage of using a more expensive box for the test?

Secondly, before you rag on the scientific value of the comparison, you might want to read the details.

The Dell server DID use SCSI. It was a Dell 1650 Poweredge server, equipped with 3x36MB drives equipped in a SCSI RAID array.

To answer your other concern, the DELL also has two 64bit/66mhz PCI slots. So you can cross that difference off your list as well.

Details on the Dell 1650 Poweredge are available at:

http://www.dell.com/us/en/bsd/products/model_pedge_2_pedge_1650.htm

thanks jase71

as a side note, 10.2 builds are much faster than that of 10.1.x Much much faster to be more precise. I am curious of how the test would fare if using 10.2 in a few weeks. Just curious. Hell, it may not benefit from it, but it may. hopefully apple finishes that up soon.

[LegendKiller]

Ok, so now if Dell DID use its scsi sub-system which should be tied to the 64/66 pci's, using 15k drives then I call the whole benchmark BS, especially considering the apple used IDE. You cannot even get anywhere NEAR 80mb/s out of IDE. I dont trust places that would post such absurd claims unless I saw the actual pic's from Sandra or whatever they used to test.

I did the reviews of the drives last week, and I have a couple more drives. Seagates X15's only get to 70mb/s, the Gen 3's only get just above that.

Furthermore, when considering IDE and throughput the Ram takes an even larger factor in the benchmarks.

Then, considering that IDE beats SCSI in this type of thing (basic throughput with few users).

Lastly, take out the ingegrated SCSI and drives and you remove about 800-1k right there. Add that to the BS 80mb/s IDE claims and you get the Dell winning.



Originally posted by jase71


First of all, Dell's pricing structure is hardly Apple's problem. If Dell can't equip a comparable server for the same dollar amount, why should they be given the advantage of using a more expensive box for the test?

Secondly, before you rag on the scientific value of the comparison, you might want to read the details.

The Dell server DID use SCSI. It was a Dell 1650 Poweredge server, equipped with 3x36MB drives equipped in a SCSI RAID array.

To answer your other concern, the DELL also has two 64bit/66mhz PCI slots. So you can cross that difference off your list as well.

Details on the Dell 1650 Poweredge are available at:

http://www.dell.com/us/en/bsd/products/model_pedge_2_pedge_1650.htm

[ribitch]

Originally posted by LegendKiller
Sure, MS doesn't write its down drivers, It depends on the IHM's to make them, and many times those co's write crappy drivers.

Sure, the price is similar now, but thats only because apple is going for market penetration, not income maximization

Sure, Apple is allowed independant channels, but they are NOT real world examples. They are unfair benchmarks created to spread FUD about PC's. A real world benchmark is comparing 2 like systems at a price point. Not buying 2 same-priced, crippling one in order to make the other win. In a standards-based market like computers you should compare like to like, but muddying this line is where Apple gets its profits.

LK


thats complete crap. this test isnt pushing apples or dells or suns. Its simply showing how the companies software compares on systems. It so happens to be apple and dell systems were similarly priced. how is a real world test 2 like systems at a same price point. the price point varies by vendor and their services. Nobody crippled one to make the other win. The company is concerned about selling their software not selling computers. if that was the case, they would sell computers on their site.

This demonstrates file access and page generations. all of which computers do in the WORKING world. most other benchmarks use quake and unreal as a standard. These dont matter in business.

common sense tells you to get the better performing system for a few hundred dollars cheaper.

[SnowSurfer]

some people buy things on price...others buy things on how they fare against other systems. so if i take a dell p3 with a geforce3 and match it up against a alienware athlonxp 2200 with a geforce4, but the athlon is stripped down so it costs the same as the dell full system, and the alienware comes out on top. does that seem fair to you?

[DoPeY5007]

Originally posted by SnowSurfer
some people buy things on price...others buy things on how they fare against other systems. so if i take a dell p3 with a geforce3 and match it up against a alienware athlonxp 2200 with a geforce4, but the athlon is stripped down so it costs the same as the dell full system, and the alienware comes out on top. does that seem fair to you? if you were comparing the two systems on cost, then yeah that is fair

[LegendKiller]

You still dont get it.

1. The results HAVE to be BS. Why? Because it has been proven HUNDREDS of times that the scsi subsystem is MUCH better and claiming an 85mb/s throughput on IDE drives is complete BS

2. When considering that the Benchies are BS and even if they aren't and SCSI somehow loses (which is NOT possible), then that would place the dell in a different level as far as servers go (more high-end database stuff with its high-disk read/write randomization that SCSI is good at). Thus, comparing the Dell server to the Apple server is stupid. Strip the Dell down WITHOUT scsi and then you get it comming below Apple. Put in the same drives and you get the same performance.


You are not seeing the economics of this. I am in finance/IT so I see both ways AND I have a scientific background so I can see it objectively based on extraneous variables. You on the other hand are just saying "It wins because they say it does". If you believe anything anybody tells you then I have some oceanside property in North Dakota that I have for sale for 10k per lot.

LK




Originally posted by ribitch


thats complete crap. this test isnt pushing apples or dells or suns. Its simply showing how the companies software compares on systems. It so happens to be apple and dell systems were similarly priced. how is a real world test 2 like systems at a same price point. the price point varies by vendor and their services. Nobody crippled one to make the other win. The company is concerned about selling their software not selling computers. if that was the case, they would sell computers on their site.

This demonstrates file access and page generations. all of which computers do in the WORKING world. most other benchmarks use quake and unreal as a standard. These dont matter in business.

common sense tells you to get the better performing system for a few hundred dollars cheaper.

[jase71]

Originally posted by LegendKiller
Ok, so now if Dell DID use its scsi sub-system which should be tied to the 64/66 pci's, using 15k drives then I call the whole benchmark BS, especially considering the apple used IDE. You cannot even get anywhere NEAR 80mb/s out of IDE. I dont trust places that would post such absurd claims unless I saw the actual pic's from Sandra or whatever they used to test.

I did the reviews of the drives last week, and I have a couple more drives. Seagates X15's only get to 70mb/s, the Gen 3's only get just above that.

Furthermore, when considering IDE and throughput the Ram takes an even larger factor in the benchmarks.

Then, considering that IDE beats SCSI in this type of thing (basic throughput with few users).

Lastly, take out the ingegrated SCSI and drives and you remove about 800-1k right there. Add that to the BS 80mb/s IDE claims and you get the Dell winning.




Ok... so first you were unhappy because the Dell didn't have SCSI and 66mhz PCI.

Then you found out the Dell HAD SCSI and 66mhz PCI, so you claim it's not fair because the Mac has IDE, and the numbers are "impossible".

However, had you bothered to actually READ the study, you'd see that the Apple used RAID as well. Can you get 80MB/s out of RAID? Use enough drives, and yes, you can... even with IDE RAID.

Now... maybe I didn't go to the #1 Psychology school in the nation... but the crappy little state university I DID attend taught me to at least read through the details of the study before I agreed or disagreed with it. It keeps me from looking like an idiot when I try to defend my stance.

[ribitch]

ok, then why is it ok for PC users to compare a powermac which uses PC133 with a DDR based windows systems? PC users are often proud of that feat, but how legendkiller is talking, the test is completely invalid.

Is the test invalid because apple uses RISC and Dell uses CISC? Intel does have a MHz lead over the G4's. 400 MHz per CPU, whicg means 800 MHz total. Thats almost a full CPU!! :D

Hell if they needed to be indentical, does that mean the cases need to be the same?

[ribitch]

Originally posted by jase71


Ok... so first you were unhappy because the Dell didn't have SCSI and 66mhz PCI.

Then you found out the Dell HAD SCSI and 66mhz PCI, so you claim it's not fair because the Mac has IDE, and the numbers are "impossible".

However, had you bothered to actually READ the study, you'd see that the Apple used RAID as well. Can you get 80MB/s out of RAID? Use enough drives, and yes, you can... even with IDE RAID.

Now... maybe I didn't go to the #1 Psychology school in the nation... but the crappy little state university I DID attend taught me to at least read through the details of the study before I agreed or disagreed with it.

as i previously satated in this thread, isnt it fun to disprove people and harass those who DONT read the entire article? I'm amazed at all the hits this thing has generated. WOW!!

[LegendKiller]

In order to reach 80mb/s with RAID you would need about 4 drives running RAID0, you dont get anywhere near double the performance when running raid0. Now, considering the failure rate of IDE drives you got another problem.

I didn't see that they were running RAID, doesn't matter much though. However, this puts it into a VERY different category. You cannot compare IDE raid results to single-disk SCSI results, again, its absurd. If I were an IT person looking for a server, I would see two totally different systems, one running tried and true scsi, and another running IDE RAID0. One has a higher price but is using ONE drive (much more reliable AND faster), and the other using multiple drives to lower is price and raise its performance.

Still you are comparing IDE to SCSI which is MUCH more expensive. Its like comparing the previous porche to a souped up honda, your little turbo honda may get you faster but it will blow up faster also.

Thus just validates further comparing apples to oranges, you guys just shot yourselves in the foot, and so did apple when any IT professional reads that.


I corrected myself on the 64/66 before.

[LegendKiller]

I never voted for either, I just merely pointed out what was wrong with the comparison.

I think that apple has made another big snafu in targeting a market with the wrong product. Sure, people will buy it when they look at the numbers, and dont care its IDE (which STILL doesn't compare to the fastest drives, and thus the benchies are still suspect). However, once those drives running in RAID0 start failing they will go down and burn.

You have to remember, the one thing that is the credo of IT is BACKUPS and RELIABILITY. You cannot get that with RAID0 unless you are running RAID10 (In which case you are wasting a lot of money and drive space), and then you sacrifice more speed. You can do it running RAID5, but that invalidates the stated numbers.

Uneven comparisons will not attract companies who are frugal and want the best system for the best money and want reliability and longevity. Thus, you will NEVER see those servers in a data warehouse, or even smaller servers. They will be reserved for lower-end duties, much what apples are relagated to now.

I like apples for many reasons, but they have had VERY bad execution and business strategy. They have targeted themselves into a corner and are trying to get out, but they will not be able to do so. Their hardware innovation is awesome, but their marketing is aweful as far as the common man, and the average IT professional goes. Until they change that they will never be a big player.


Originally posted by ribitch
ok, then why is it ok for PC users to compare a powermac which uses PC133 with a DDR based windows systems? PC users are often proud of that feat, but how legendkiller is talking, the test is completely invalid.

Is the test invalid because apple uses RISC and Dell uses CISC? Intel does have a MHz lead over the G4's. 400 MHz per CPU, whicg means 800 MHz total. Thats almost a full CPU!! :D

Hell if they needed to be indentical, does that mean the cases need to be the same?

[jase71]

Originally posted by LegendKiller

I didn't see that they were running RAID, doesn't matter much though. However, this puts it into a VERY different category. You cannot compare IDE raid results to single-disk SCSI results, again, its absurd.

Thus just validates further comparing apples to oranges, you guys just shot yourselves in the foot, and so did apple when any IT professional reads that.




Ok... now you're not even reading the posts in this thread, much less the study.

The Dell WASN'T single disk SCSI.

The Dell was SCSI RAID as well.

BOTH systems used RAID.

That's what, the THIRD time you've "shot yourself in the foot"?

[NuTs62]

haven't read the rest of the thread.. not sure if this has been brought up.. but isn't photoshop designed using Apple or Macs..? its kinda a lopsided comparison.. try comparing a program designed more for PC's.. and the PC should out bench it.

[LegendKiller]

Ok, I looked to see where you were getting that stuff, I didn't even see the tabs at the top of the overview page for the system specs, so sorry if I am not perfect.

I still do not see SCSI raid being beat by IDE raid, especially 36gig X15's in a basic throughput benchmark. That is VERY fishy, especially considering the high-cpu overhead for IDE (software/hardware) raid.

There IS something wrong with that. Furthermore, it doesn't say what type of RAID. If the dell was using RAID 5 (3 drive config) and the Apple was using RAID 10 (4 drive config, 2 stripped with other 2 mirror, or 4-striped config) the RAID 10/0 would win hands down. Again, that puts the apple into a TOTALLY different class NOT based on price but features. Find a dell with RAID0 IDE's and you would get a good price/performance comparison.

While I AM being corrected, the corrections are only proving my point.

LK



Originally posted by jase71


Ok... now you're not even reading the posts in this thread, much less the study.

The Dell WASN'T single disk SCSI.

The Dell was SCSI RAID as well.

BOTH systems used RAID.

That's what, the THIRD time you've "shot yourself in the foot"?

[ribitch]

Originally posted by NuTs62
haven't read the rest of the thread.. not sure if this has been brought up.. but isn't photoshop designed using Apple or Macs..? its kinda a lopsided comparison.. try comparing a program designed more for PC's.. and the PC should out bench it.

ha, i knew somebody would post this. This ISNT RUNNING photoshop on the server. The client is running photoshop, and openeing a file stored on the server. The server is serving the file. Thus the throughput is measured. Therefore any optimization isnt aiding in the results, since it all caries on how fast the server can send the files.

Photoshop was designed for macs, and I believe both OS's have performance enhancers and optimizations now.

I wont harass you though, since you admitted you havent read the thread.

[jase71]

Originally posted by LegendKiller

I still do not see SCSI raid being beat by IDE raid, especially 36gig X15's in a basic throughput benchmark. That is VERY fishy, especially considering the high-cpu overhead for IDE (software/hardware) raid.

There IS something wrong with that.


Perhaps. Or perhaps OSX is just a more efficient OS when it comes to high volume file serving. With it's Unix core, it's got a history of running on big iron.

Or perhaps they were running different flavors of RAID. It doesn't look like they tell us what type of arrays they were running.

But unless you can find some legitimate reason the numbers are unreasonable, or some past history of Xinet fudging the numbers, I'm going to trust them.

To make sure your 80MB/s numbers were valid, I took a look at a few reviews on www.storagereview.com - and I found
that 80MB/s is an attainable number with a 3 IDE drive array. I see no reason why I should doubt Apple can reach those numbers on a brand new system, when storagereviews numbers came from an IDE RAID card running on a slower, 33mhz bus PCI slot.




Furthermore, it doesn't say what type of RAID. If the dell was using RAID 5 (3 drive config) and the Apple was using RAID 10 (4 drive config, 2 stripped with other 2 mirror, or 4-striped config) the RAID 10/0 would win hands down. Again, that puts the apple into a TOTALLY different class NOT based on price but features. Find a dell with RAID0 IDE's and you would get a good price/performance comparison.

If, if, if...

All you've done is speculate.

Unless you've got some hard facts to present...




While I AM being corrected, the corrections are only proving my point.


Not quite. :rolleyes: But keep trying. :P

The irony is, I'm not an Apple guy. I own three pcs... A linux machine, a Win2K video-editing machine, and a Win2K box for my wife. No macs.

In fact, I work for a company that builds and sells pcs and pc-based servers.

But if Apple builds a great server for the money, more power to them. Competition only helps the buyers.

[ribitch]

Originally posted by jase71


Perhaps. Or perhaps OSX is just a more efficient OS when it comes to high volume file serving. With it's Unix core, it's got a history of running on big iron.

Or perhaps they were running different flavors of RAID. It doesn't look like they tell us what type of arrays they were running.

But unless you can find some legitimate reason the numbers are unreasonable, or some past history of Xinet fudging the numbers, I'm going to trust them.

To make sure your 80MB/s numbers were valid, I took a look at a few reviews on www.storagereview.com - and I found
that 80MB/s is an attainable number with a 3 IDE drive array. I see no reason why I should doubt Apple can reach those numbers on a brand new system, when storagereviews numbers came from an IDE RAID card running on a slower, 33mhz bus PCI slot.




If, if, if...

All you've done is speculate.

Unless you've got some hard facts to present...




Not quite. :rolleyes: But keep trying. :P

The irony is, I'm not an Apple guy. I own three pcs... A linux machine, a Win2K video-editing machine, and a Win2K box for my wife. No macs.

In fact, I work for a company that builds and sells pcs and pc-based servers.

But if Apple builds a great server for the money, more power to them. Competition only helps the buyers.

damn, he shoots and he scores!!

where has darkfury been to support dell? I am shocked he hasnt been active in this.

Anyways, the results are there. The xserve has just started to ship, so i am sure we will see some more benchmarks soon. legendkiller, you can fight all day, but you havent shown anything solid yet. buy what you want to buy, support what you want to support. let the numbers convince those who are unsure. plain and simple. I am shocked this thread has drawn so much attention.

[LegendKiller]

I do/not agree that OSX could be more efficient in file serving due to its Unix heritage. While it DOES have that heritage it has been morphed into something completely different.

The if's are introduced by their own inability to publish the details

The 3-ware card might even be the same card they are running, it is ATA-100. Furthermore, at that level, PCI 64/66 might not matter (A bit of a back-up on previous statements but that was when I was not sure if the 1650 or the mac even both had 64/66). If it does, then it might make a difference. 80mb/s is attainable, but I do not see how SCSI raid could perform WORSE in this case, unless it was using a different type of raid, one which IT professionals would not care to run (0), which again, puts it into a different price/performance/feature class and invalidates the comparison made.


LK



Originally posted by jase71


Perhaps. Or perhaps OSX is just a more efficient OS when it comes to high volume file serving. With it's Unix core, it's got a history of running on big iron.

Or perhaps they were running different flavors of RAID. It doesn't look like they tell us what type of arrays they were running.

But unless you can find some legitimate reason the numbers are unreasonable, or some past history of Xinet fudging the numbers, I'm going to trust them.

To make sure your 80MB/s numbers were valid, I took a look at a few reviews on www.storagereview.com - and I found
that 80MB/s is an attainable number with a 3 IDE drive array. I see no reason why I should doubt Apple can reach those numbers on a brand new system, when storagereviews numbers came from an IDE RAID card running on a slower, 33mhz bus PCI slot.




If, if, if...

All you've done is speculate.

Unless you've got some hard facts to present...




Not quite. :rolleyes: But keep trying. :P

The irony is, I'm not an Apple guy. I own three pcs... A linux machine, a Win2K video-editing machine, and a Win2K box for my wife. No macs.

In fact, I work for a company that builds and sells pcs and pc-based servers.

But if Apple builds a great server for the money, more power to them. Competition only helps the buyers.

[LegendKiller]

I would LOVE to get numbers on this, just for my own self-gratification. I dont care who would win, but I would just be interested in creating a valid scientific experiment and publishing results that would be useful.

As I said, perhaps we can try and find some way to get Apple to give Gotapex a Xserve and dell (or home-grow) a wintel server and compare them. It would give GA a huge presence and garner some good attention. I will talk with Darth about that.

My next review is going to be a Gen 2.5 Seagate Cheetah X15...


LK


Originally posted by ribitch


damn, he shoots and he scores!!

where has darkfury been to support dell? I am shocked he hasnt been active in this.

Anyways, the results are there. The xserve has just started to ship, so i am sure we will see some more benchmarks soon. legendkiller, you can fight all day, but you havent shown anything solid yet. buy what you want to buy, support what you want to support. let the numbers convince those who are unsure. plain and simple. I am shocked this thread has drawn so much attention.

[jase71]

Originally posted by LegendKiller
I do/not agree that OSX could be more efficient in file serving due to its Unix heritage. While it DOES have that heritage it has been morphed into something completely different.

It's not really all that different at all. You can run Xwindows on OSX. You can easily port linux/unix apps over to OSX.

In fact, Darwin, the open source core of OSX, is actually installable on PCs.

Apple just slapped their own GUI and a lot of frosting on top of a Unix core. The core itself has changed very little from a Unix system. OSX is unix, through and through.

The morphing was done on the GUI and the eye candy, not on the core of the OS.



The if's are introduced by their own inability to publish the details

The 3-ware card might even be the same card they are running, it is ATA-100. Furthermore, at that level, PCI 64/66 might not matter (A bit of a back-up on previous statements but that was when I was not sure if the 1650 or the mac even both had 64/66). If it does, then it might make a difference. 80mb/s is attainable, but I do not see how SCSI raid could perform WORSE in this case, unless it was using a different type of raid, one which IT professionals would not care to run (0), which again, puts it into a different price/performance/feature class and invalidates the comparison made.

Maybe true.

But unless we KNOW there was a difference in RAID Arrays, we can't really blame the Dell's performance problems on it.

What if the RAID arrays were the same? (which is actually the most likely scenario) THEN what might be the cause? Maybe, just maybe, the Apple really is an excellent system.

I would assume, that because Xinet has a good reputation, and because they tested a variety of systems from Sun to SGI to Apple to Dell, that they kept things as similar as possible, including running the same types of RAID arrays.

If you can show me a case of Xinet behaving irresponsibly in the past, or of fudging numbers, or show me that the RAID arrays were in fact different, then I'd be more inclined to take your point seriously.

But claiming that they MUST be different simply because you disagree with the numbers... well... that's not persuasive enough for me.

[SnowSurfer]

wow we can speculate speculate speculate but u know what (apple defenders) u cant really say anything unless u are 100% of the configurations and u give links. ooo wow the apple server can beat dell p3's omg thats amazing. no it really isnt. i want to see who is going to purchase one of these things because i cant see the practical implication in a pc world. explain how you would implement this? yes the server is cheaper than a dell but it comes out more expensive. if u think about it how much money would u have to pay to train people? more than that of the price difference.

[DoPeY5007]

Originally posted by SnowSurfer
i want to see who is oing to purchase one of these things because i cant see the practical implication in a pc world. it isn't a PC world everywhere... most design company only use mac's, and one place I worked @ used macs for the art people and PCs for all the data entry people... So mac server will be a better option for some people, it just depends on what you do and what you need

[jase71]

Originally posted by SnowSurfer
explain how you would implement this? yes the server is cheaper than a dell but it comes out more expensive. if u think about it how much money would u have to pay to train people? more than that of the price difference.

The core of the OS is still Unix.

Most large companies have a fair number of Unix servers. And they have a fair number of Unix admins.

OSX would let them roll out smaller servers that are easily manageable.

Unix admins, with a minumum of training, could admin the OSX boxes. After all, all their CLI experience will apply to OSX as well.

If anything, it would make it easier on the server guys, since OSX and Unix will play together better than Windows and Unix.

[Speedfreak]

Originally posted by ribitch


i dont think its so empty...
theres an emmy (http://www.apple.com/firewire/)

5 recent IDEA awards (http://www.businessweek.com/magazine/content/02_27/b3790101.htm) (Industrial Design Excellence)

Powerschool has a 2002 Codie (http://www.siia.net/codies2002/winners.htm)

If you want to get into software feats, apple dominates withGenetech Blast. Theres also the Seti@home factor, the iPod, and powerbook.

seems to me apple has several recent awards and envies. sorry

No, you wouldn't see the personal preference, would you?

[ribitch]

Originally posted by SnowSurfer
wow we can speculate speculate speculate but u know what (apple defenders) u cant really say anything unless u are 100% of the configurations and u give links. ooo wow the apple server can beat dell p3's omg thats amazing. no it really isnt. i want to see who is going to purchase one of these things because i cant see the practical implication in a pc world. explain how you would implement this? yes the server is cheaper than a dell but it comes out more expensive. if u think about it how much money would u have to pay to train people? more than that of the price difference.

let me see...

easily managaed and maintained, OS stability and robusteness, apache claiming OS X is one off teir fastest impolementations, fast affordable processors, largest unix distribution..

hmm, maybe education, small companies large enough to need rack mounted solution, webhosts, design firms, production companies, higher education, independent directors... there are more than you think

Oh, and with apple recent aquisition of many large audio/video software companies, and statements stating windows support will cease after certain dates, I can see apple making one hell of a Xserve based workstation for audio/video production.

Apple has been very active lately. Several products have been a hit (powerbook, ibook, ipod, imac) and this may be the next hit for apple. Many of you are doubters know, but wait 6 months, and you will probably be shocked at how much market share apple will capture. This may not do it all, but the recent "real people" campaign has drawn attention. Its partially because the people are flacky, but made you pay attention. Apple stores are popping up all over the place, and about 50% of systems bought are from PC converts or people looking to play with a mac. iPod is considered one of the ultimate mp3 players, and powerbook is viewed as one of the ultimate laptops. People buy what they like, and what they want. This has caused plenty of positive press fro apple.

in a time of tech layoffs, apple is still hiring. Apple has turned profits, is once again an admired company. They apparently are doing something right....

[Apex]

Originally posted by ribitch

in a time of tech layoffs, apple is still hiring. Apple has turned profits, is once again an admired company. They apparently are doing something right....

Apple marketshare climbed roughly 5.4% Q1/2002, but issued lowered earnings warnings with revenues down about 10 percent due to weaker sales to consumers and advertising and publishing businesses. Problems were especially strong in Europe and Japan. Apple's US marketshare is now roughly 3.7%, up from 3.4% last year. This is good news, as Apple has been on a historical decline (6.7% in 1996, 4.1% in 1997 in the US).

http://www.pegasus3d.com/macmarketshare.gif














Meanwhile, Dell jumped 19%.

[ribitch]

interesting, but this says it best

http://www.oreillynet.com/pub/wlg/591

apple is here to stay. Tell me, what protion of the market is sun? or SGI?

so dell rose. still wont make me like them. I may accept them more if they were to lead the PC industry and start dropping all floppies and legacy ports, but that wont happen until intel gives the go ahead.

Apple is an inovator. we all know its a fact. Apple had firewire back when they adopted USB. At the time, the windows world was amazed at plug and play with USB while apple had bigger and better technologies standard on the powermac. PC users still laugh at apple for dropping the floppy, yet they never seem to use it. Instead tehy have CDR-W or ZIP.

This forum is becomming more and more mac friendly. even though the about us section makes a crack at macs, we are here, and we can exist together happily (that sounds really lame)

[jase71]

Originally posted by ribitch
I may accept them more if they were to lead the PC industry and start dropping all floppies and legacy ports, but that wont happen until intel gives the go ahead.

There are already motherboards out there, like the Abit AT7, that are dropping PS2 ports, serial ports, parallel ports, and the like. The AT7 has audio in/out, USB, and Firewire ports. Other than IDE, that's about it.

I've seen cheap econobox systems built around similar motherboards, albeit without the firewire...

I haven't had a floppy drive in my main pc in a couple of years... it's easy enough to disable the controller in BIOS.

Legacy stuff is going away, even in PCs...

Unfortunately, PCs carry the burden of having to support a larger number of corporate and business users, where legacy stuff hangs on for an obscene amount of time. I have to lecture users every week about keeping critical files on network drives, rather than on floppy disks. My client uses an insane number of label printers that only come with serial hookups... and all the pcs use NT4, which means no USB. So getting rid of legacy hardware, at least for them, isn't even an option for at least several years yet. So someone had better make some of those legacy boards for at least a few more years...

[ribitch]

not everyone will always be happy. People will eventually be forced to upgrade. The sooner the better.

people need to realize legacy can actually be a crippling factor to a system

[SnowSurfer]

ribitch i think u posted this topic just to get a reaction from people and to try and make people look small. are trying to say only smart people use macs? hmm thats kinda funny. the whole world must be stupid. except the mac users and mac people. but u know its kinda funny this topic evolved into a mac vs pc. how big is the mac user base compared to pc. there is only 1 company that makes mac's compared to the dozens that make pcs. maybe that is mac's downfall. when was the last time u heard of someone overclocking a mac or watercooling a mac or slapping a brand new case on a mac hmm....maybe pcs are more customizable and user friendly after all. or maybe u are a mac recruitor...who knows. yes i have gone into the mac store and played with everything and stuff and yes i would buy a powerbook for the sole purpose of connecting my digital video camera to it but u know what. that wouldnt be cost effective and thats what companys want cost effectiveness. how would u go about selling one of these servers to a company. not in the way u have been talking in this fourm. u would be kissing their bum to sell them one if your commision was on the line. so try and sell me one in a civil manner. i want to see :)

[DREDD]

Well, whoever said that Macs can't play games should read the article...was that a games comparison? I know you didn't refer to the article, but common...we all know the PC is for games!

Now...to be clear...my 2.63Ghz P4 runs Photoshop great, but my father has a dual 1ghz G4 mac with a gig of RAM at work and it runs Photoshop up to 50% faster in some tests. That's a beating for my Pc. Course the mac is made for work not for games. my PC was built for games and let's not compare FPS between these systems haha

[DREDD]

actually snowsurfer I've seen super cooled macs with refrigerators and such things hooked to them and clocked real high and custom cases and such. the mac market is an enthusiest market. It's real expensive to be a mac user, but what you get is a machine that always works 100% of the time and that's no lie. I've had my current G3 for about 4 years and it still works like i just bought it. You don't buy a mac to play games like i said before.

[verve247]

Snowsurfer wants Ribitch to kiss his bum? :hmm:

[Ladogaboy]

Sorry I haven't been able to reply in a while... you know, work got in the way...

Anyway, after reading all the previous posts, I do think LK has a legitimate point (along the same lines as I was making originally)...why weren't they much more open about the hardware configurations? I would think, in a review like this, you would want to include EVERY detail of how the hardware was run, setup, and used.

Also, I have a couple of other questions: If price was the motivating factor for comparing the Apple with the Dell, why did they include the other, higher-priced servers? And if they were going to do that, why didn't they use the $7K Apple versus the $7K Dell?

[Showtime]

Apples rule!
Go Apple
my 2 cents
-jel:halo:

[Jeffbx]

Wow - the fur is flying on this topic!

I think it's great that Apple is exploring this market, but I'm confused about a couple things -

- Why rack mount? Many small business don't have rack mount capabilities. Why not at least offer both rack & desktop?
- Why so small? Sure, a 1u device is great for server farms, but I would think that Apple would want to get their foot in the door before going for ultra high density. Switch to 2U or 3U, add room for some additional drives & space to cool them, and they would be much better off.
- Why IDE? The independent controllers is a very cool idea, but think of the throughput of that setup with SCSI instead! :gle: I understand there would be a cost difference, but for a server, I wouldn't even consider IDE. The cost difference would be worth it.

It's a great first attempt on Apple's part, tho.

As far as the comparion with Dell - who cares? At that price point, it's a matter of OS preference, not cost, that will make the decision.

[ribitch]

i never said PC users are stupid or dumb. They are often uninformed or misinformed about other platforms, and threads liek this are attempts to inform them. Many PC users are stubborn in their ways and wont listen, while others will

Games on the mac are available, but not as widely available. NVidia's new programming language is supposed to help that change, but who knows when we will see results. Apple has equivilant video cards with the PC world now (GF4 series, GF 2 series, ATI Radeon series), but a game manufacturer has yet produced something to fully utilize the video abilities. Sure we can only pull 180 fps in Q3 running in classic mode on OS X. The game is running on an emultor. Thats like running Q3 on VPC and expecting it top match unemulated mode. Its clear that PC users have the gaming edge.

Upgradabilty i spossible on a mac. It happens more often than one may think. You can buy new video cards, audio cards, scsi cards, Ultra ATA cards, CPU upgrades, DV encoding cards, NIC's, and many more internal options. People dont realize that apple has a great audio susbsystem in their systems, great networking abilities, and plenty of CPU power. This means these are harder to come by. Theres a reason SB lives! dont sell as well in macs as they do with PC's. Its because their audio system is already superior to that of a PC. If you wanted to invest money in audio, you could get one hell of a soundsystem, but most people dont need a professional audio I/O system.

Overclackability - Apples are often overclocked. Maybe not as often as a PC, but people do it. Water cooled, liquid nitrogen, refrg., peltier, you name it and it can be done. Often easier on a mac as well due to their case design. Hell, people even have websites telling how to overclock apple laptops. Its all out their, you just arent looking in the right places.

cases are available for macs, and people do change them. Most people in the mac world like their cases as they are. They are designed with excellent airflow, easy access, and easy expandability. Most people are not going to put their iMac in a new case, but some people do. With the powermacs, wires do not restrict airflow. Apple has put a great amount of detail into tucking away the cables and allowing the maximum amount of airflow.

software - Most peole tend to use on average 4-5 apps on a pc, and 6-7 on a mac. Thats a fact, not something i pulled out of my ass. Most people do not need winzip, powerzip, pkunzip, and many more zip programs. You simply use 1 app to do that task. "well PC's have more apps in general" Its about quality not quantity. PC's may have many more good apps, but they have by far more many poor apps. Number of apps isnt the question. What types of apps do you need and what ones are actually usefull is.

I believe the higher priced servers were used to show what the lower end servers do when compared to a high end server. Some printhouses often use sun servers for their storage. Is this why? Only the testing company knows. Why didnt they include full specs? No clue. I bet they will release them to you if you asked. They gave model numbers, CPU speeds, HD interface, HD size, and RAM size. For most people thats sufficeint. They all used RAID. They didnt tell what method they used, which could be for serverl reasons. What if their software us designed to use only on 1 type of RAID. "All tests were performed on a single RAID/DISK on a SINGLE controller" taken from
link (http://xinet.com/aboutxinet/pdf.library/configurations.pdf/xinet.benchmark.config.2002.pdf) Does this mean that a raid controller with only 1 disk on it was used? I dont know. Ask them. The Xserve and Dell both used the internal RAID (Xserve IDE RAID, Dell Int SCSI Raid) controllers. So no aftermarket parts were used

[ribitch]

Originally posted by Jeffbx
Wow - the fur is flying on this topic!

I think it's great that Apple is exploring this market, but I'm confused about a couple things -

- Why rack mount? Many small business don't have rack mount capabilities. Why not at least offer both rack & desktop?
- Why so small? Sure, a 1u device is great for server farms, but I would think that Apple would want to get their foot in the door before going for ultra high density. Switch to 2U or 3U, add room for some additional drives & space to cool them, and they would be much better off.
- Why IDE? The independent controllers is a very cool idea, but think of the throughput of that setup with SCSI instead! :gle: I understand there would be a cost difference, but for a server, I wouldn't even consider IDE. The cost difference would be worth it.

It's a great first attempt on Apple's part, tho.

As far as the comparion with Dell - who cares? At that price point, it's a matter of OS preference, not cost, that will make the decision.

They did offer powermac based servers, and they can be configured like that. Customers wanted a Rack based solution though, so apple listenmed.

1U is great because you can fill a rack and run it in a beuwolf configuration. Macs are used in biotech/math/video for raw CPU power and this configuration allows plenty of it in a confined space.

Software RAID mirroring or striping is also an option, although Apple said a companion RAID hardware product with 2Gb fibre channel will be available by the end of the year
taken from maccentral (http://maccentral.macworld.com/news/0205/14.xserveannc.php)

also

And Apple trailed a dedicated storage product Xserve RAID: a dual 2Gigabit Fibre Channel array with room for 14 bays in a 3U rack. That'll be out later in the year, the server next month. from the register (http://www.theregister.co.uk/content/53/25278.html)

[Jeffbx]

Originally posted by ribitch



also

quote:
--------------------------------------------------------------------------------
And Apple trailed a dedicated storage product Xserve RAID: a dual 2Gigabit Fibre Channel array with room for 14 bays in a 3U rack. That'll be out later in the year, the server next month.
--------------------------------------------------------------------------------

from the register (http://www.theregister.co.uk/content/53/25278.html)

That's the one they should have started with....

[ribitch]

Originally posted by Jeffbx


That's the one they should have started with....

i believe this is just a storage device. I could be mistaken though since very little was given on it.

It would be nice to see that end up being a high end server though. Maybe a quad processor?

[Ladogaboy]

Wow... PC users only use 4-5 apps on average. I don't buy that in the least... that means that there are people out there that turn on their PC just to play solitaire and then shut down again. I would have to count to be sure, but I personally use well over 20 apps regularly. :shrug:

[LegendKiller]

Originally posted by Ladogaboy
Wow... PC users only use 4-5 apps on average. I don't buy that in the least... that means that there are people out there that turn on their PC just to play solitaire and then shut down again. I would have to count to be sure, but I personally use well over 20 apps regularly. :shrug:

Yeah, I found that tremendously hilarious. Why? Because this whole thread the mac users have been saying "show me numbers on why you think that these benchmarks are incorrect. Show me information on IF the raid's are different, all you have are more "IF'S"".

Thats great, and then they go in and say "I KNOW that the average PC user uses XX amount of programs". How do you know the average PC user? Considering a massive number of PC's reside in the work force and a majority of those are multi-taskers, I HIGHLY doubt the average is 4-5. That just implies that PC users are stupid simpletons who either dont know how to use their PC's more efficiently, or that Joe-6-pack only checks his email on his e-machine.

Lets see, take my mom. She uses Photoshop, Eudora, IE, about 6 other digital picture programs, not to mention all of the games she plays, and she is a 56 year old woman. I would bet she uses around 20 programs.

Me? At my job I usually have 6 programs open at once, lotus, IE, word, excel, PPT, Snagit, crystal ball, and many others.

At least I can back up my numbers with logic, even if it is quite fuzzy, but at least I have some background information and my questions about the raid setups are valid.

If the raid setups were the same, then how were they set up? Are you telling me that the raid in the Dell was 3-drives and the raid in the Mac was 3 drives? What about the 4th drive in the Mac? Was it just set aside? Or did the Mac have a RAID5 with 4 drives while the dell had it only with 3. If that were true, then the Mac might win just because of the extra I/O's from the other drive.

Do we know that the Gigabit controllers were the same? Were they plugged into the same type of PCI slot?

how about the client PC computers...they would make a HUGE difference in the opening speeds. Were those standardized?

You say the Xinet has an "excellant record". How do you know what kind of record they have? Everybody thought Enron, Qwest, Waste Management, Aruther Andersen (a company of over 80 years no less), Martha Stewart, Worldcom, et al also had an "excellant record". Everybody always has one...until they get caught or they release "grey data" which Xinet has done.

What it comes down to, is that you nor a PC zealot will give each system a fair playing field, nor the users a fair shake at equality. You claim price-point advantage, but that is only had by using cheaper drives in a pump-up configuration that reduces reliability, something that carries a lower value in IT admins eyes. PC users claim they have the advantage in programs, but then many crash, again carrying lower validity in the enterprise world.

You all look at it through your own "beer goggles", but do not realize that what you are saying carries little to no logic or proof. Numbers are thrown down, statistics are quoted (which even factual statistics numbers can be bent more than a lead wire), but then when it comes down to leveling the playing field and actually testing the TRUE subsystem of a computer, both sides cry foul.

Why? Because your innate inability to see logic from an unbiased outlook.


LK

[jase71]

Just out of curiosity, LK...

You've tried to tear down this study.

But you haven't provided any logical solutions.

How exactly would YOU compare an Apple server and a Dell server, for example? How would YOU benchmark them? What hardware would YOU pick?

Show us how you'd set up an "unbiased" comparison.

[ribitch]

I am currently looking for the report on average number of applications used. This isnt talking about apps like minesweeper or AIM, its talking about things like lotus, office, adobe apps, etc. I will post it when I find it. I have stumbled across many interesting articles about cost of ownership, comparable systems, and things of that nature, so its taking longer to find. I would post these, but I dont want to sidetrack this from the original topic.

I found one interesting article on apple firsts that were brought over to the PC. Word and excel were originally on the mac (i never knew that), and the givens like the 3 1/2 floppy, internal cd-rom, mouse, etc.

[LegendKiller]

In psychology, a study is not valid unless you have a control group and a test group. The control group is a baseline while the test group is what you give your actual indepedant variables. For example, you give the control group a placebo and the test group the actual drug and measure the results.

However, in order to have a valid experiment, you must eliminate every possible extraneous variable, within reason. For example, if I wanted to do a study on how one hard drive compares to another, I cannot use 2 different test systems. I would have to use the same system, the same controller, and the same formatting on each drive.

However, in this case, it was not done. They were not truly testing a Mac v dell server. They were testing 4-drive IDE raid+DDR v 3-drive scsi raid+PC133SDram. To any IT professional who is going to buy servers, this comparison is rediculous. The main price-advantage that Mac has in this case is the fact that 3-36gig Scsi drive+integrated controller = about 1500 in price. 3-120gig WD's+integrated raid controller= about 500 dollars. Thats 1000 dollar difference. Not to mention the realiability premium that the IDE drives lack...which is accounted for by price, but also buy brand-awareness.

To truly test a mac v pc-based server one would need to eliminate every possible extraneous variable and conform those to the price-point presented.

how would I do that? I would take that 4500 mac server price and then find, or configure a dell to be similar. Take a base-line dell P4 with a faster type memory, RDRAM or DDR, and then use IDE raid. Or better yet, take the mac and configure it with SCSI 3-drive raid.

Or, just max out the mac specs and then see what I could find by dell for around the same price point. A maxed out MAc will not have such an odd-ide config to it.

However, even these would not be totally equal. The only way to do that would be to config the mac to the specific price-point, and then BUILD a pc (since there are no PC-based servers that will be configed the same) to match it.

What benchmarks? Well, there would have to be a custom built benchmark, or some 3rd party program like Xinet's program.

What would be the same on both computers?

Ram-exact same manufacturer of chips, same nanosecond rating

hard drives- same manufacturer, model, and interface

SCSI controller- Same manufacturer, model, and interface and cables

Gigabit cards- same manufacturer, put into same type of slot, if its a shared-IRQ slot then make sure both either share or do not share an IRQ with another card.

Thus, there would be only 3 variables left, CPU, motherboard sub-systems, and OS efficiency, all of which would ultimately determine which system was better.

Even more beneficial would be the ACTUAL price-performance ratio. Why? Because if you price other variable EXACTLY the same (hdd's, controllers, etc) then the only price variences would be the actual subsystems that are in question.

Why will we NEVER see this? Because people like direct comparisons. Furthermore, when you want a server you want one with the 5-year 24x7 warrenty, which a self-built server DOESN'T have. So comparisons like this will never happen unless you have a truly unbiased and very dedicated website.

I would literally piss my pants if I, or even GA as a whole were able to do this. It would probably be the biggest thing on the net, a TRUE battle of OS/hardware. Who would win? Who knows, both architectures have their advantages. Would people be happy? Heck no, no matter who won or lost, even the winners wouldn't be happy...and that is when you KNOW you have a good test.

As my GF's father says, and I truly believe, you know you have done the best thing when EVERYBODY is pissed at you.

LK

[jase71]

Originally posted by LegendKiller
However, in order to have a valid experiment, you must eliminate every possible extraneous variable, within reason. For example, if I wanted to do a study on how one hard drive compares to another, I cannot use 2 different test systems. I would have to use the same system, the same controller, and the same formatting on each drive.

When car magazines compare cars, do they equip them with the same tires? The same brands of air filters? Oil filters? Do they make sure they run the same brand of motor oil?

No... they test them as they came from the factory. They test them as the manufacturer provided them.

Removing extraneous variables is nice... but what you're testing here is how well the servers perform, at a price point, as they come configured from the manufacturer.

These are not scientific studies. They're product comparisons. This is Consumer Reports, not JAMA.



However, in this case, it was not done. They were not truly testing a Mac v dell server. They were testing 4-drive IDE raid+DDR v 3-drive scsi raid+PC133SDram. To any IT professional who is going to buy servers, this comparison is rediculous.

Sure they were. An IT Professional looking to purchase a server isn't going to pull all the drives out of the systems and configure them on the same controller. They're not going to rip out the included memory and make sure the same brand was used.

They're going to use the product how it was shipped from the manufacturer.

Each change you make introduces variables into the test that the end user will not duplicate. Then you're not testing the systems, you're testing the quality of the changes made TO the systems. Did you do a better job on the changes you made to the Apple than you did to the Dell?

When testing entirely different platforms, it's most fair to leave the systems as stock as possible. Otherwise you run into a situation like Netcraft did a while back, where they tested a completely tweaked Microsoft server against a bare-install Linux machine.



The main price-advantage that Mac has in this case is the fact that 3-36gig Scsi drive+integrated controller = about 1500 in price. 3-120gig WD's+integrated raid controller= about 500 dollars. Thats 1000 dollar difference. Not to mention the realiability premium that the IDE drives lack...which is accounted for by price, but also buy brand-awareness.

Yes, but these are BENCHMARK tests... not quality and reliability tests. Nowhere in the study does it discuss MTBF. The study doesn't even attempt to cover it.

So to bash the study because IDE drives were used on some servers, and SCSI on other is unfair. Xinet didn't even try and address that. This study was simply about benchmarks at a price point. They used the servers that were provided, that matched a price point.

IS MTBF an issue for IT Professionals? Sure, of course it is. But it's one of many factors, and it's a factor that this survey didn't even pretend to cover. It's material for a different study.



how would I do that? I would take that 4500 mac server price and then find, or configure a dell to be similar. Take a base-line dell P4 with a faster type memory, RDRAM or DDR, and then use IDE raid. Or better yet, take the mac and configure it with SCSI 3-drive raid.

You CAN'T build a Dell dual-P4 server for less than $5,000. And that's with NO controller, and NO hard drives.

And if you purchase an IDE controller and hard drives and add them on your own, then you're testing a heavily modified Dell server, not a machine as shipped from the manufacturer.

What do you do if you can't build equivalent systems for close to the same dollar amount? What if a fair hardware matchup means an grossly unfair dollar amount?

Is it fair to compare a $20,000 Ford to a $40,000 Chevy, even if the basic specs are the same?





What would be the same on both computers?

Ram-exact same manufacturer of chips, same nanosecond rating

hard drives- same manufacturer, model, and interface

SCSI controller- Same manufacturer, model, and interface and cables

Gigabit cards- same manufacturer, put into same type of slot, if its a shared-IRQ slot then make sure both either share or do not share an IRQ with another card.


To do that, you'll have to modify the systems heavily. One of the systems will need the factory-installed RAM removed.

What if the same hard drives, controller, and network cards aren't available? Which system do you modify to make them match? Are you showing bias if you make the Dell change hard drives to match the Apple? Does the study have ANY value if you make each server use drives that aren't available for the readers to purchase from the manufacturer?

Not to mention, what you'll end up with are systems that bear very little resemblance to the ones that shipped from the manufacturer.

Is it fair then to post benchmarks on those systems, considering the hardware is dramatically different than what the customer will be getting when he orders one?

Your benchmarks would be worthless, because the systems they came off of wouldn't resemble the ones for sale even remotely.




Why will we NEVER see this? Because people like direct comparisons. Furthermore, when you want a server you want one with the 5-year 24x7 warrenty, which a self-built server DOESN'T have. So comparisons like this will never happen unless you have a truly unbiased and very dedicated website.


Comparisons like this CAN'T happen. Because you'll never match hardware specs and prices closely enough to be completely unbiased.
The difference is especially pronounced when you're trying to compare two completely different platforms.

All you can really do is set a price point, and let each manufacturer provide what he feels is his best value at that price point. Then you benchmark them, and see who gives you more bang for the buck at that price point.

Any changes you make to the hardware detract from the value of the comparison, because you'll make the products dissimilar to the ones the readers will potentially buy. They don't want to read about what you can turn an Apple or Dell server INTO. They want to compare what you GET when you buy one.

[ribitch]

Originally posted by jase71


When car magazines compare cars, do they equip them with the same tires? The same brands of air filters? Oil filters? Do they make sure they run the same brand of motor oil?

No... they test them as they came from the factory. They test them as the manufacturer provided them.

Removing extraneous variables is nice... but what you're testing here is how well the servers perform, at a price point, as they come configured from the manufacturer.

These are not scientific studies. They're product comparisons. This is Consumer Reports, not JAMA.



Sure they were. An IT Professional looking to purchase a server isn't going to pull all the drives out of the systems and configure them on the same controller. They're not going to rip out the included memory and make sure the same brand was used.

They're going to use the product how it was shipped from the manufacturer.

Each change you make introduces variables into the test that the end user will not duplicate. Then you're not testing the systems, you're testing the quality of the changes made TO the systems. Did you do a better job on the changes you made to the Apple than you did to the Dell?

When testing entirely different platforms, it's most fair to leave the systems as stock as possible. Otherwise you run into a situation like Netcraft did a while back, where they tested a completely tweaked Microsoft server against a bare-install Linux machine.

[B]

Yes, but these are BENCHMARK tests... not quality and reliability tests. Nowhere in the study does it discuss MTBF. The study doesn't even attempt to cover it.

So to bash the study because IDE drives were used on some servers, and SCSI on other is unfair. Xinet didn't even try and address that. This study was simply about benchmarks at a price point. They used the servers that were provided, that matched a price point.

IS MTBF an issue for IT Professionals? Sure, of course it is. But it's one of many factors, and it's a factor that this survey didn't even pretend to cover. It's material for a different study.



You CAN'T build a Dell dual-P4 server for less than $5,000. And that's with NO controller, and NO hard drives.

And if you purchase an IDE controller and hard drives and add them on your own, then you're testing a heavily modified Dell server, not a machine as shipped from the manufacturer.

What do you do if you can't build equivalent systems for close to the same dollar amount? What if a fair hardware matchup means an grossly unfair dollar amount?

Is it fair to compare a $20,000 Ford to a $40,000 Chevy, even if the basic specs are the same?





To do that, you'll have to modify the systems heavily. One of the systems will need the factory-installed RAM removed.

What if the same hard drives, controller, and network cards aren't available? Which system do you modify to make them match? Are you showing bias if you make the Dell change hard drives to match the Apple? Does the study have ANY value if you make each server use drives that aren't available for the readers to purchase from the manufacturer?

Not to mention, what you'll end up with are systems that bear very little resemblance to the ones that shipped from the manufacturer.

Is it fair then to post benchmarks on those systems, considering the hardware is dramatically different than what the customer will be getting when he orders one?

Your benchmarks would be worthless, because the systems they came off of wouldn't resemble the ones for sale even remotely.




Comparisons like this CAN'T happen. Because you'll never match hardware specs and prices closely enough to be completely unbiased.
The difference is especially pronounced when you're trying to compare two completely different platforms.

All you can really do is set a price point, and let each manufacturer provide what he feels is his best value at that price point. Then you benchmark them, and see who gives you more bang for the buck at that price point.

Any changes you make to the hardware detract from the value of the comparison, because you'll make the products dissimilar to the ones the readers will potentially buy. They don't want to read about what you can turn an Apple or Dell server INTO. They want to compare what you GET when you buy one.

Wow, you are full of suprises. I am shocked once again.

I work for a university. Wehn we order a server, we order it based on service from the company, the servers performance, the stability of teh system, and most importantly price. We do have people that are stuck on their ways. So we will have people who say dell is the only way to go or compaq or whoever.

LK, your views on how to benchmark these are so irrelevant. You are saying almost every aspect of teh system needs to be changed. If you changed anymore, they would be the same system, and a benchmark would be worthless. You benchmark a system on how the manufacturer ships it, or by comparing itself to itself with an upgrade.



how about the client PC computers...they would make a HUGE difference in the opening speeds. Were those standardized?

If you read the report using HTML or PDF, you would know that these are standardizeds. Once again, read the story first then post

[LegendKiller]

Ok, I am going to go down the line here...


1. Testing a car and testing a computer are VERY different. A car takes an extreme amount of VERY unique charactaristics and tests them together. Not one part on a car from each manufacturer is the same. A car is not modular but computers are, even between manufacturers.

2. Different people look at different things. As an IT professional myself, and somebody well versed in finance I look at both sides of a computer. If I just took a strait-line comparison of each componant on those 2 systems I would immmediately discount said amounts of money for not only value-addedness but also reliability. This is the way its done everywhere which has a budget. You dont only look at the base sticker price, but you look at the features you get for that price. Thus, in order to compare 2 systems you need to configure it for the same price.

Now, you may say "well, I get the same or better PEFORMANCE for the same price. However, any professional IT person, CIO, or techie will realize WHERE that performance came from and what cost it will have in the future.

Lets take this scenario as an example.

My GF's father is a CFO. He was approached 9 months ago, by his boss (CEO) to do a systems upgrade on the accounting software, EDI hardware, and the POS terminals of 9 stores in geographically unique locations. The requirements were that the systems must last 5 or more years, be able to report revenue figures in less than 3 days after the year end close, and to be able to provide semi-accurate revenue figures every day.

Now, this required EVERYTHING to be new, servers, POS terminals, phone exchange hookups, and software to match.

I actually sat in with my GF's dad and compared the 4 different companies systems. They went through each one, hardware, software, and compared that to price.

It came down to "If system A has this at $xxx dollars then how does this compare to system B that has this at $xxx".

Now, if the CIO were handed the dell and then the Mac systems he would make the following observations.

1. "The dell is from a company who KNOWS servers"
2. "The dell has SCSI raid, but the Mac's transfer rate is faster. IDE drives are cheaper and this is where I would get a benefit. More for less".
3. However, the CFO would say "What would be the return on that speed? Would the financial benefit equal the downsides to using this type of system"
4. CIO would say "IDE RAID may be faster, but it also prone to failure, furthermore it doesn't scale as well as SCSI"
5. The CFO would say "If that were true, then can we get a Mac server with SCSI raid?"
6. The CIO would say "Yes, we can, but instead of being 300 cheaper, we would end up spending 1000 more on the Mac"
7. The CFO would say "Ahhh, so in order to get SCSI we have to pay a premium?"
8. CIO "Yes, we do. So if we found a PC server that could take IDE raid, then we would have a server that would be MUCH cheaper than the mac".
9. CEO "BUT, if we did that, what would the performance be like? Would the mac win by an even larger margin?"
10. CIO "That is doubtful for serveral reasons. 1. 4-drive IDE raid is going to be faster than 3-drive SCSI raid, giving a performance boost to whichever system runs it. Thus, the mac would be slower without it and the PC would be faster with it. While this might not totally normalize the performance levels the systems would be more apt to perform in the same manner. 2. If we normalized the IDE-raid, then we would be able to also configure a server with higher-end CPU's and memory, furhter taking the advantage away from the Mac server".

10. CFO "So its not a question of which system is faster, but one of which system has what parts that MAKE it faster, and those parts are not really Mac or PC specific but generic standards"?

11. CIO" That is correct, we will be able to find a system that can match the Mac, perhaps we can even work with dell to get this type of system, especially if we order enough they might even give us a discount"

12. CFO "what if we work with apple?

13. "We could get a discount there also"

14. "Ok, get with both of them, see if you can give me figures on which system can do what when they are similarly configured, when that is done we can do an ROI analysis on each system, whichever has the better returns over its life, including maintenance (especially since IDE raid will have more maintenance costs), tech support, and downtime (again, IDE=more), then we will make a decision.



Now, you may not think this will happen in the real world, but i KNOW it does. I was there for this same exact thing. Apple ruins it credibility with the true movers when they say "we are better at XX price" but then they do not really show WHY they are better, but when the person realizes why, they see they have been hoodwinked into even THINKING that.

YOu may think your pretty hot stuff sitting on your mac, but your the reason why apple is small and your thinking is the reason why it will remain so. By not realizing that your small beans and being elitist you only put apple into a niche and ensure that it will not reach its full potential.


LK

[LegendKiller]

I hate to say it, but university spending is VASTLY different from the business world. Again, while you may think universities spend a ton of money, they often spend it unwisely, and even that amount is nothing compared to what real businesses spend. Until apple realizes that they cannot play games with numbers/parts/and prices they will be nothing to the high-end enterprice community.


LK



Originally posted by ribitch


Wow, you are full of suprises. I am shocked once again.

I work for a university. Wehn we order a server, we order it based on service from the company, the servers performance, the stability of teh system, and most importantly price. We do have people that are stuck on their ways. So we will have people who say dell is the only way to go or compaq or whoever.

LK, your views on how to benchmark these are so irrelevant. You are saying almost every aspect of teh system needs to be changed. If you changed anymore, they would be the same system, and a benchmark would be worthless. You benchmark a system on how the manufacturer ships it, or by comparing itself to itself with an upgrade.


If you read the report using HTML or PDF, you would know that these are standardizeds. Once again, read the story first then post

[LegendKiller]

Originally posted by jase71


When car magazines compare cars, do they equip them with the same tires? The same brands of air filters? Oil filters? Do they make sure they run the same brand of motor oil?

No... they test them as they came from the factory. They test them as the manufacturer provided them.

Removing extraneous variables is nice... but what you're testing here is how well the servers perform, at a price point, as they come configured from the manufacturer.

These are not scientific studies. They're product comparisons. This is Consumer Reports, not JAMA.


I thought of something to add in this.


When CR does a comparison of a variety of cars, they list EXACTLY what is being done and how. They list the price of each car AND the configuration. Thus, the person can do an accurate comparison between price AND features.

Furthermore, they do not put a cheap sports sedan into the budget sedan market segment. Why? because even a cheap sports sedan will beat a budget sedan, even if the sports sedan IS cheaper. Why? Because its configured differently. Furthermore, a person looking for a budget sedan would not even be interested in what the cheap sports sedan does because they are looking for a CERTAIN MARKET! They dont care if the cheaper sports sedan has a 4-cyl with dual turbo's, all they care is about their 6-cyl engine that runs quieter, but does not care if its top speed is 20mph less.

Thus, each car is aimed at a different market and are NOT a valid comparison. In order to do a valid comparison you should compare cheap sports sedan to cheap sports sedan.

What can I get for 20k or less? I can buy a Camry, a used Toyota Spider, a pontiac Grand Am, A chevy cavalier, or I could even get a cheap minivan. Do you see why comparing just on price/performance can yield so many different results? Each one is targeted at a different market, some for the foreign/cheap/butsporty others for the domestic/cheap/butpractical, and yet another for foreign/cheap/used/sporty. Then, add in the fact that you could buy a product that instead of being a 2003, but a 2000 but NEW, then you add in a whole new category...domestic/cheap/old/butpractical. Maybe I could find a SOHO dual overhead cam Tuarus then I would get domestic/cheap/old/butveryfast and then compare it to a foreign/cheap/new/butsupedupwith2turbo's.

Again, two totally different markets. Buyers realize that, but when you cutting and pasting this and saying "MAC WINS MAC WINS, BOO TO PC's" you make a VERY stupid statement.


LK

[ribitch]

me think of myself as hot stuff, not quite.

if a company will sit down and analyze the cost of a system, its maintenance, its uptime/downtime, its rated average age, and the total cost over that average rated age, then I'm sorry to tell you this, but there would be many more macs sitting on desks today

According to the August 2001 report conducted by PCWorld, Dell has a 3.0 year average age. The highest of any system besides apples 3.9 years. Of all major PC makers, the average age was 2.4 years. When you calculate the cost of maintenance per year per PC (on average is between 40-50 hours per system) and 11 hours per mac, the lost productivity because of the downtime (1 work week per year per PC), the cost of the repairs, plu sthe initial cost of the system, you get what variou research firms have discovered. According to the latest gartner report (its a pay to see it type report), apple systems cost about 400 dollars less per system to run in a business enviroment. So you sit here saying that all this is calculated by business before they purchase a system? WRONG. IT support people see apple as less machines to support, which means less work, which means fewer jobs. many IT support people that have experience with a mac will agree that they are by far easier to work with, fix, and keep up.

The low amount of maintenance has actually hurt apple because of this aspect. Thats why they are hovering around 5% of teh market share (i dont know where apex pulled his fgures from. no source was given. 5% is what all the analyst use as the going figure)

Now if we were to get into power consumption, the cost of a PC will also rise. A PC consumes much more power than a mac does. spread teh difference over 50 systems in an office at 7 cents per hour and you will see costs rise. Why dont businesses calcutlate these as well?

The new iMac is rated at 130 Watts per hour power consumption under full load. 7W in sleep. My old 1.2 GHz AMD system was over 300W per hour. Thats over double!!!

Once again, its sheer performance obtained per the dollar. In this test, an early test for the xserve, the apple performs better at less expense than the dell. Deny it as much as you want, but the facts are there

[jase71]

Originally posted by LegendKiller

1. Testing a car and testing a computer are VERY different. A car takes an extreme amount of VERY unique charactaristics and tests them together. Not one part on a car from each manufacturer is the same. A car is not modular but computers are, even between manufacturers.


Parts are only partially modular between the Apple and the Dell system. You can't interchangeably swap anything and everything.

And you still haven't addressed the fact that by changing so many parts, you completely negate the value of the study, because you've changed the systems into something that clearly doesn't relate to what the customer will be purchasing.





2...14 <snipped for brevity>



All nice points, mostly true, and completely irrelevant to the topic.

Remember, this began as a simple set of benchmarks between several servers, including a Dell and an Apple.

The study listed hardware, and the benchmarks.

It did NOT attempt to address cost of implementation, cost of upgrades, service, or any of the other things you've tried to drag into the discussion.

Those are all perfectly valid considerations when considering which server your business is going to buy. But in a simple benchmark test, they have absolutely no bearing on the outcome. None of those factors change anything in Xinet's study. ROI and downtime do not affect IO performance of a RAID array.

Had Xinet attempted to suggest which server was better, and to persuade us which one we should purchase, all the things you mention would come into play. Considerations like downtime would weigh heavily, as would ROI, and modularity of the hardware.

But they didn't. All they did was post numbers.

So squiggle and squirm as you will, the factors you mentioned simply don't matter, at least for the topic at hand.




YOu may think your pretty hot stuff sitting on your mac, but your the reason why apple is small and your thinking is the reason why it will remain so. By not realizing that your small beans and being elitist you only put apple into a niche and ensure that it will not reach its full potential.


LK

Reread the earlier posts. I'm not the Mac guy. I don't own a Mac. I'm the guy who works for the company that makes PC-based servers.

Whose bias is showing?

[LegendKiller]

Again, you are throwing out numbers without providing me links.

1. From personal experience Macs are easier to fix when there is a problem. However, the diagnosis takes MUCH longer, and the problems keep showing up. I do not see how even get close to saying those figures. Again, I need hard numbers WITH LINKS. As you see, I do not list numbers like that, I list logic because its much easier to deal with as long as you are talking to a logical person.

Also, how do they figure "average life"? Is it the fact that most mac users keep their mac because it "does what they want", well, if thats true then you can keep an Imac (targeted audience is the stupid email user) and it can last well past the quantum computing era, as long as it can hook up to 100gb ethernet through firewire/usb. Then its average life is 100 years...big whoop.

Again, unfair comparisons. PC's are workhorses, not artwork.

2. A desktop and a server are 2 VERY different things. A desktop goes down, 1 person cannot work. A server goes down, and 100 people cannot work. Take IDE, put it into a Mac server, and when that sucker goes down, 100 people sit around with their thumbs up their butts...business's wont like that.

3. As far as your wattage numbers, they can be complete BS. The fact that AMD lists a 300w PS as its manditory number is actually because 300w PS's produce more power, whether you have a good or bad one. However, if you have a bad 135w it wont run.

Take for example an Enermax 431w PS, at any given time its sustained power output is about HALF of its 431. HOWEVER, its a constant half and the power spikes are much less. Then, you take your run of the mill craptastic buy-a-case-with-a-PSU PSU, you get a 235-300w, it will be lucky if it even pushes 1/4 of the volts through the rails that the 431 does. Its all the QUALITY sent out and not the QUANTITY "consumed".

Furthermore, one VERy large advantage for macs is that the PSU they want you to have is the one you have. AMD cannot control that, thats why they "recommend" one. Did you know the average dell PSU is around 180w? I saw a compaq 1.33ghz Athlon with a 200w. Pretty surprising, especially since that sucker wont even get anywhere near 200w consumption.

Dont even try to BS me on numbers like that. Especially considering the difference in 135 and 300w is NOWHERE near 7c per hour. People have figured it out (I dont keep track of power rates) and I have heard (since you want to pull numbers from nowhere) that keeping your PC on 24/7 WITH the monitor on will cost you aroun 30c per MONTH. The monitor is the biggest power sucker there, so take that down to .12 per month....not quite 7c per day.

Furthermore, figure in that what you say isn't even close to true and your advantage is GONE.

[ribitch]

see page 10 of this
http://www.hubster.com/apple/whymac/idc-tco-white-paper.pdf

in education, mac users use more apps than pc users. The paper (by IDC sponsored by apple) discusses costs of computers in schools. It mentions business in some areas, but not all areas. I am still looking for the average number of apps used in a non education enviroment

education
mac 12
PC 9.7

they further go on to say that the more apps used, the more you learn. Makes sense, you are not only learning that software, but also the subject that it might be teaching.

Yes, i know apple paid for this study to be done, but IDC is independent from apple.

[jase71]

Originally posted by LegendKiller


When CR does a comparison of a variety of cars, they list EXACTLY what is being done and how. They list the price of each car AND the configuration. Thus, the person can do an accurate comparison between price AND features.

And, if they forget to include some information, do you automatically assume the numbers are completely skewed, wrong, and unfair?

Or do you trust, based on their past reputation, that the numbers are accurate, until you find solid evidence that contradicts them?




Furthermore, they do not put a cheap sports sedan into the budget sedan market segment. Why? because even a cheap sports sedan will beat a budget sedan, even if the sports sedan IS cheaper. Why? Because its configured differently.

Neither do they take each car, and try and make sure the horsepower is exactly the same by modding one car so the specs match the other.
They don't pull parts off of one until they weigh the same.

Yet that's what you'd have had Xinet do with the servers... swap parts until the cars were the same, removing parts from the manufacturer, and replacing them with parts NOT from the manufacturer.
How on EARTH would that be representative of the items for sale to the public?

The magazines compare the cars as they come from the manufacturer, not modified so the specs are identical.



What can I get for 20k or less? I can buy a Camry, a used Toyota Spider, a pontiac Grand Am, A chevy cavalier, or I could even get a cheap minivan. Do you see why comparing just on price/performance can yield so many different results?

So... you're suggesting the Dell server is a minivan. And the Apple server is a slightly sporty economy car?

Please, enlighten me. What class of server was the Dell, and what class of server was the Apple? Considering they were both tested as file servers, how was the Dell classed inappropriately compared to the Apple? The Dell even had the advantage of SCSI.

In the framework of the study, benchmarks, what ~$4500 Dell server (unmodified, purchasable from Dell) SHOULD they have used?

[LegendKiller]

What parts aren't modular? Mac and PC have IDE, PCI, USB, AGP, ATX, 5.25 and 3.5", and DIMMS, that pretty much takes care of everything BUT the CPU and motherboard, which is truly what should be compared.

No, changing the parts doesn't negate the study, it makes it so we are comparing like to like. Ok, you keep saying that we are comparing performance for "off the shelf" stuff. However, you do not realize that the IT professional (which the article is geared towards) is NOT going to do that. He WILL look at the underlying attributes of each one and he WILL compare the price/peformance of each. He WILL know HOW the mac one and HOW that can change when compared to the Dell.

I am not biased towards either one. I am biased towards people who post stupid comments like "We won because we are cheaper and better using cheaper but faster parts, and we beat a more expensive system...but....the other system has BETTER parts in it and more EXPENSIVE parts in it".

You CANNOT compare a 20 year old rusted Lamborghini with 4 out of 12 pistons working to a brand new honda civic with "performance" enhancers. They may be almost the same price, but with 2 hands tied behind the lambo's back you cannot expect it to win.

I cannot stand FUD and I hate even more people who spread it. Its like my friend who is a Linux Zealot. I once asked him WHY he likes Linux. He said "because I hate microsoft", thats EXACTLY what linux companies DONT want. They want you to like linux because linux is BETTER, not because its "Not something else".

Mac people hate PC's because they are "wintel", and they are "slower" but when they test out a PC they find the best performing Mac and then take a "off the shelf" PC of their own choosing. Its complete BS. But its the ONLY way apple has survived, by muddying the waters they convince the uneducated that they are better.


LK





Originally posted by jase71


Parts are only partially modular between the Apple and the Dell system. You can't interchangeably swap anything and everything.

And you still haven't addressed the fact that by changing so many parts, you completely negate the value of the study, because you've changed the systems into something that clearly doesn't relate to what the customer will be purchasing.


[B]

All nice points, mostly true, and completely irrelevant to the topic.

Remember, this began as a simple set of benchmarks between several servers, including a Dell and an Apple.

The study listed hardware, and the benchmarks.

It did NOT attempt to address cost of implementation, cost of upgrades, service, or any of the other things you've tried to drag into the discussion.

Those are all perfectly valid considerations when considering which server your business is going to buy. But in a simple benchmark test, they have absolutely no bearing on the outcome. None of those factors change anything in Xinet's study. ROI and downtime do not affect IO performance of a RAID array.

Had Xinet attempted to suggest which server was better, and to persuade us which one we should purchase, all the things you mention would come into play. Considerations like downtime would weigh heavily, as would ROI, and modularity of the hardware.

But they didn't. All they did was post numbers.

So squiggle and squirm as you will, the factors you mentioned simply don't matter, at least for the topic at hand.


[B]

Reread the earlier posts. I'm not the Mac guy. I don't own a Mac. I'm the guy who works for the company that makes PC-based servers.

Whose bias is showing?

[LegendKiller]

Originally posted by ribitch
see page 10 of this
http://www.hubster.com/apple/whymac/idc-tco-white-paper.pdf

in education, mac users use more apps than pc users. The paper (by IDC sponsored by apple) discusses costs of computers in schools. It mentions business in some areas, but not all areas. I am still looking for the average number of apps used in a non education enviroment

education
mac 12
PC 9.7

they further go on to say that the more apps used, the more you learn. Makes sense, you are not only learning that software, but also the subject that it might be teaching.

Yes, i know apple paid for this study to be done, but IDC is independent from apple.

Do you know how of total computer sales education comprises? Less than 2% probably. Give me what the AVERAGE user uses and not some half-baked numbers of a VERY small population and then that will apply to macs AND computers in general.

Again, your numbers do nothing but put Mac where it belongs, in a very small dark corner until it can see the light

[ribitch]

i correct myself once again

it was pc magazine, not pc world

dell is only 2.0 years compared to apples 3.9 years

i would post the gartner groups findings, but its a pay to read it report.

life expectancy
http://www.pcmag.com/article2/0,4149,87189,00.asp

theres some more facts. They even came from a PC magazine, not a mac magazine.

As far as problems reoccuring on a mac, or taking longer to fix, have you ever considered that you may have no clue on what you are doing? That could be your problem right there. If you dont know how to fix it, and find a half ass way to fix it, it will come back eventually. Hell, home users can often fix there mac with no problems at all. Its not that hard if you educated yourself on it or simply followed some directions on fixing the problem

[LegendKiller]

So your one of those maroons who kept investing in Enron despite warnings from people for the last 3 months? How about Rambus? Want some oceanside property in North Dakota?

I suspect everybody who doesn't publish the full story. Moronic investors lost their own money through their own fault by investing in Enron. Did you invest 10m of your 10.0001m life savings in Enron? Well, you deserve your 100 bucks because I have no pity for you. A good investor is a smart investor, is a cautious investor is an EXTREMELY suspecting investor...ask warren buffet if he had his analysis go over enrons statements with a fine tooth comb...then ask how many shares Berkshire had...

As I said before, everybody looks good until they get caught.




Originally posted by jase71


And, if they forget to include some information, do you automatically assume the numbers are completely skewed, wrong, and unfair?

Or do you trust, based on their past reputation, that the numbers are accurate, until you find solid evidence that contradicts them?

of course they dont level the horsepower, but they do put each car into its own category. An IDE server is NOT the same as a SCSI based one...even if the price and performance is same or better






Neither do they take each car, and try and make sure the horsepower is exactly the same by modding one car so the specs match the other.
They don't pull parts off of one until they weigh the same.

Yet that's what you'd have had Xinet do with the servers... swap parts until the cars were the same, removing parts from the manufacturer, and replacing them with parts NOT from the manufacturer.
How on EARTH would that be representative of the items for sale to the public?

The magazines compare the cars as they come from the manufacturer, not modified so the specs are identical.

Again, the only true comparison would be a server from dell with IDE raid. Why do you refuse to acknolwege the only performance and price advantage the Mac has is from IDE? If you take that away all you would get is an overpriced desktop Mac...wow, just like everything else they sell. They have to CHEAT to get ahead.

The TRUE comparison would be to have an XServe configured with 3x36gig SCSI in raid configuration. At least then you would get comparible prices if not also performance. But guess what...ohhh crap, they DONT SELL ONE!!! Why? Because they want to look good on paper with prices, AND they want to keep performance by using 4xdrive raid. WOW, they are getting cake and EATING IT TOO!!!

Take their precious "cheap 4500 server", take away their 4xRAID, add 3x SCSI drives and you get a server thats not only 1000 bucks more than the dell, but also much closer in performance.


LK



So... you're suggesting the Dell server is a minivan. And the Apple server is a slightly sporty economy car?

Please, enlighten me. What class of server was the Dell, and what class of server was the Apple? Considering they were both tested as file servers, how was the Dell classed inappropriately compared to the Apple? The Dell even had the advantage of SCSI.

In the framework of the study, benchmarks, what ~$4500 Dell server (unmodified, purchasable from Dell) SHOULD they have used?

[LegendKiller]

Originally posted by ribitch
i correct myself once again

it was pc magazine, not pc world

dell is only 2.0 years compared to apples 3.9 years

i would post the gartner groups findings, but its a pay to read it report.

life expectancy
http://www.pcmag.com/article2/0,4149,87189,00.asp

theres some more facts. They even came from a PC magazine, not a mac magazine.

As far as problems reoccuring on a mac, or taking longer to fix, have you ever considered that you may have no clue on what you are doing? That could be your problem right there. If you dont know how to fix it, and find a half ass way to fix it, it will come back eventually. Hell, home users can often fix there mac with no problems at all. Its not that hard if you educated yourself on it or simply followed some directions on fixing the problem

Well, considering I have a masters degree and I work in IT/Finance and have 6 years of sys admin at a big-10 school with a 30/70 mac/pc split and a trouble-shooting database that had 300 problem/solutions after I left...and considering about 70% of those were from macs, I would say that I had a pretty good clue.


LK

[jase71]

Originally posted by LegendKiller
What parts aren't modular? Mac and PC have IDE, PCI, USB, AGP, ATX, 5.25 and 3.5", and DIMMS, that pretty much takes care of everything BUT the CPU and motherboard, which is truly what should be compared.

Just because the systems share a slot format, like AGP or ATX, doesn't mean the parts are interchangeable.

For example, most AGP video cards will not work if you simply plug them into an AGP Mac. Same goes with PCI video cards. And vice versa. Most Mac AGP and PCI video cards will not work if you simply plug them in to a PC AGP or PCI slot. Even if you have the appropriate drivers. It doesn't matter.

The same applies to IDE and SCSI cards. Just because the form factor is the same does NOT mean the hardware is interchangeable between systems.



No, changing the parts doesn't negate the study, it makes it so we are comparing like to like. Ok, you keep saying that we are comparing performance for "off the shelf" stuff.

I see. So if I take a nice Dell system, and stick in a Rage II video card, 16MB of memory, and a 2 gig hard drive, and publish benchmarks, those are valid? Even if I equip the Apple with equivalent hardware? Those are fair benchmarks?

I don't think so!

What use is it to a reader to read a study about a server with a configuration he can't possibly buy from the manufacturer? How can he use that information?

Add to that, if you modify all your benchmarks for these two systems, and you review a new system next month, do you equip THAT one the same as these? Or do you set new standards for each review?

If you modify and tweak each system that comes through, then none of your reviews have any continuity. Benchmarks from one review have no connection to another. You can't compare them, even if they were run identically.



However, you do not realize that the IT professional (which the article is geared towards) is NOT going to do that. He WILL look at the underlying attributes of each one and he WILL compare the price/peformance of each. He WILL know HOW the mac one and HOW that can change when compared to the Dell.

He won't know jack, because the numbers you would have given him would have been so skewed he wouldn't have a baseline to make any judgements from. He won't know how the server would have performed out of the box. He won't know how fast the default config was. He won't know if the original drives were fast enough for him.

All he'll know is the benchmarks that your heavily modified system acheived. But that tells him NOTHING about the original product.




Mac people hate PC's because they are "wintel", and they are "slower" but when they test out a PC they find the best performing Mac and then take a "off the shelf" PC of their own choosing. Its complete BS. But its the ONLY way apple has survived, by muddying the waters they convince the uneducated that they are better.

I like pcs better than Macs. I don't own any macs.

But when I see a system perform well, especially considering it's price, then I'm perfectly satisfied to be happy for it. I don't need to tear it down simply because it's not the kind of system I like best.

Apple made a nice server? Good for them! I'm happy for 'em. It's good for the market. It's good for Apple. It's good for anyone who would consider buying one.

Why belittle that just because I don't really want to buy one myself?
Is it a bad product just because I don't like it? No...

I just don't understand the fanboy mentality that one thing has to be the best, and the other has to suck. Apple made a nice server. Dell has always made nice servers. At this price point, the Apple might outperform the Dell. That's nice. The Dell might have other advantages that some people would prefer. That's nice.

Why does someone always have to be the bad guy?

[ribitch]

Originally posted by LegendKiller


Well, considering I have a masters degree and I work in IT/Finance and have 6 years of sys admin at a big-10 school with a 30/70 mac/pc split and a trouble-shooting database that had 300 problem/solutions after I left...and considering about 70% of those were from macs, I would say that I had a pretty good clue.
LK

i'm not doubting your experiences, i am doubting your knowledge. There is a difference. Quite franly, i dont give a rats ass that you were a systems admin at big 10 school, or have a masters, or whatever you want to brag about.

So you working on a database, thats what a knowledge base is. Pat yourself on the back.

considering the data given, and the studies done, the mac is cheaper in the long run. look at the system prices. say you spend 2500 for a nicely loaded dell, then you pay 3500 for a nicely loaded apple. split the 2500 over 2 years and you get 1250 per year. Split 3500 over 3.9 years and you get 897.43. Higher initial cost, longer overall value. You have a degree in finance, so you should be able to see that.

If we wanted to get into specifics, an older apple system also will have a higher resale value than that of dell. So your investment will yeild a much higher return.

[LegendKiller]

I am not saying one or the other has to suck. I am saying that if you ARE going to say one is losing you should NOT do it in a way which hides information or presents the products to 2 totally different markets.


I hate to tell you this, but a Seagate HDD will work in any computer made, as long as it has the correct cable/card and the standard 12/5 voltage molex connector, and that has been standard for over 15 years.

AGP? Another standard, as long as an Nvidia chipset has drivers then no matter what card it is, it will work. AGP 1.1 is AGP 1.1, thats why its called AGP, its a standard. They are all the same voltage. However, you cannot take an AGP pro card which REQUIRES an AGP-pro slot and put it into an AGP regular slot and expect it to work.


Furthermore, PCI 1.1 is PCI 1.1 as long as there are drivers for that card for OS compatibility then the card will work. Its another STANDARD.

In order to compare systems in the same MARKET/PRICE/PERFORMANCE you have to use similar or the same parts. Cross-comparing markets and claiming you win is stupid and the reason why businesses will never take apple servers seriously.

Why is there no SCSI in Xserves? Because they KNOW they would get the crap beat out of them in both performance AND price.

Are Xserves good in their own niche? Heck yeah, a decend server with IDE raid for a good price, thats awesome. Would I trust it? Heck no, not with 4-IDE drives in RAID, might as well trust a kid with a gun.


LK

[jase71]

Originally posted by LegendKiller
So your one of those maroons who kept investing in Enron despite warnings from people for the last 3 months? How about Rambus? Want some oceanside property in North Dakota?

Enron was obviously in trouble and full of crap long before they went down. The signs were there. Even a moron could see they were headed in the wrong direction when they were making HUGE dollars off the California energy crisis. It was going to come back to bite them.

Rambus was no different.

WorldCom has been in hot water for a while now.

Very seldom do good companies go bad without giving some warning.



I suspect everybody who doesn't publish the full story.

...

As I said before, everybody looks good until they get caught.


In that case, why on earth should we trust anything YOU have to say?

I guess we can't believe ANYTHING we read. Gosh, they didn't post the lot number of the CPU they used in testing. I didn't see the serial number of the floppy drive listed. Or even the brand!
What week of the year were those hard drives built?

I've never seen ALL the details even on reviews here on G|A. Here's an idea... send us links to some of YOUR reviews. Wanna bet you didn't leave at least SOME detail out of one of your reviews? Should we completely disbelieve you then? Or should we trust you until you give us a reason not to?

Thanks, but I don't like living my life in a paranoid haze. I read as much as I can, from as many sources as I can. And I'll take each of them seriously until they give me a reason not to.

Otherwise I'll end up wearing a tinfoil helmet, screaming about the government reading my thoughts, and the latest review of the Pentium5 being nothing but megahype from a biased pack of lying hyenas.

[ribitch]

back to standardized hardware

the firmaware for AGP and PCI video and PCI drive controllers are often different between platforms. thye optimize the firmware for that platform. Sure, you can get a radeon 7500 for PC, and flash it with a mac bios. Sometimes it works, but not in all cases. (that has actually been done. some cards it works on, others it doesnt)

RAM on apple can be used on PC's, but some PC ram WILL not work on Apples. Apple maintains high standards for ram used, since this aids in system stability. Many mac users were caught off gaurd by buying standard bargain basement ram for low costs, and upodating the firmware. The new firmware forced apple sspecified ram guidelines, and disabled the bad ram. Bad customer relations for doing so? Maybe. But apple fixed many stability issues that these customers complained about.

Try getting a SMC usb NIC to work on a mac. Its a USB device, but since the mac has built on NIC's, the USB isnt needed, so neither apple or SMC supports it. Same with the 3Com ones and belkin ones. Because its a universal socket, doesnt mean it will always work.

[LegendKiller]

Again, what uses are you taliing about? What defines an "average life span"?

According to all finance people I know, the average lifespan of a computer is 3 years, and thats what they use to depreciate them, at the end of those 3 years they consider them scrap, no matter what the computer.

What defines a life span? The speed of the programs being run. What defines what is a good speed? The user of the computer. What affects the speed perception? For one, your own home computer and what yo udo with it, how much money you spend on it. The reverse can be true also. You buy a new PC at work and your home seems slow and you want to get a new one. Since PC's are cheaper you think its ok so you do it. Your 3 year investment now turns into a 1 year resale because of perception and the speed at which PC's evolve.
The same cannot be said of Macs...which have just broken the GHZ barrier...and only that through pipeline lenghtening.



Heck, right now I am on a 5 year old P2-450 with a old ATI rage video card and 256mb of ram. Wow, 5 years. I asked how much these things cost new...1,600 bucks. 320 bucks per year...not bad.




Originally posted by ribitch


i'm not doubting your experiences, i am doubting your knowledge. There is a difference. Quite franly, i dont give a rats ass that you were a systems admin at big 10 school, or have a masters, or whatever you want to brag about.

So you working on a database, thats what a knowledge base is. Pat yourself on the back.

considering the data given, and the studies done, the mac is cheaper in the long run. look at the system prices. say you spend 2500 for a nicely loaded dell, then you pay 3500 for a nicely loaded apple. split the 2500 over 2 years and you get 1250 per year. Split 3500 over 3.9 years and you get 897.43. Higher initial cost, longer overall value. You have a degree in finance, so you should be able to see that.

If we wanted to get into specifics, an older apple system also will have a higher resale value than that of dell. So your investment will yeild a much higher return.

[jase71]

Originally posted by LegendKiller


AGP? Another standard, as long as an Nvidia chipset has drivers then no matter what card it is, it will work. AGP 1.1 is AGP 1.1, thats why its called AGP, its a standard. They are all the same voltage. However, you cannot take an AGP pro card which REQUIRES an AGP-pro slot and put it into an AGP regular slot and expect it to work.


Furthermore, PCI 1.1 is PCI 1.1 as long as there are drivers for that card for OS compatibility then the card will work. Its another STANDARD.

How many links do you want to illustrate my point?

Wander out tonight, and buy yourself a Mac Radeon. Take it home, and plug it in your pc. Wanna lay odds on whether you get a video signal or not? Pick any of half a dozen other Mac video cards. The majority won't work in your pc, even though they will plug into the slot just fine...

Repeat with Mac IDE and SCSI controllers.

A few of them you can get to work by flashing the video bios, or flash rom of the card to the PC version. Most of them, even that won't help.

I can provide you all the links you want, if you don't believe me.

[LegendKiller]

The major point I was making with the standards was that you can find more or less the EXACT same thing for both computers. You can always find an Adaptec U160 SCSI card for either system, so why not use it to level the playing field, same thing with video cards, a GF4 chipset will perform more or less the same on both systems.

Ribitch, you make an excellant point with the USB and such, but again, its the standards in hardware that I was pointing out, not firmware and minor configuration differences.

Another reason why mac wanted "mac only" ram was to keep the prices high and eliminate competition. I am very cynical about this.

My reviews are of the 2 seagate HDD's. One thing that I did was treated both drives the EXACT same way, same controller and everything. Only variables left are the drives themselves.

[ribitch]

Originally posted by LegendKiller
I am not saying one or the other has to suck. I am saying that if you ARE going to say one is losing you should NOT do it in a way which hides information or presents the products to 2 totally different markets.


I hate to tell you this, but a Seagate HDD will work in any computer made, as long as it has the correct cable/card and the standard 12/5 voltage molex connector, and that has been standard for over 15 years.

AGP? Another standard, as long as an Nvidia chipset has drivers then no matter what card it is, it will work. AGP 1.1 is AGP 1.1, thats why its called AGP, its a standard. They are all the same voltage. However, you cannot take an AGP pro card which REQUIRES an AGP-pro slot and put it into an AGP regular slot and expect it to work.


Furthermore, PCI 1.1 is PCI 1.1 as long as there are drivers for that card for OS compatibility then the card will work. Its another STANDARD.

In order to compare systems in the same MARKET/PRICE/PERFORMANCE you have to use similar or the same parts. Cross-comparing markets and claiming you win is stupid and the reason why businesses will never take apple servers seriously.

Why is there no SCSI in Xserves? Because they KNOW they would get the crap beat out of them in both performance AND price.

Are Xserves good in their own niche? Heck yeah, a decend server with IDE raid for a good price, thats awesome. Would I trust it? Heck no, not with 4-IDE drives in RAID, might as well trust a kid with a gun.
LK

the socket is only half of the picture. the firmware is what matters. plain and simple.

harddrives are a given they work on all IDE equiped systems.



And Apple trailed a dedicated storage product Xserve RAID: a dual 2Gigabit Fibre Channel array with room for 14 bays in a 3U rack. That'll be out later in the year, the server next month.


taken from the register (see previous post)


Fibre Channel is the next storage interface. Fibre Channel has been adopted by the major computer systems and storage manufacturers as the next technology for enterprise storage. It eliminates distance, bandwidth, scalability, and reliability issues of SCSI.
taken from http://www.fibrechannel.com/technology/overview.html

apple has a solution for scsi

[LegendKiller]

And I suppose they will take a 15 drive RAID-5 FC Cheetah 10k XServe and compare it to a 10 drive RAID-5 U160 Cheetah 15k Dell.

1. Its FC and the dell isn't
2. It has 50% more drives with parallel I/O's
3. Sure, the Dell has faster/more expensive drives, but it can't beat FC+15drive raid
4. The dell will be more expensive in that config since 15k drives are super expensive, even up against 10k's.


Why would they do this? because they love unfair comparisons with higher-priced systems.


I will never believe anything Jobs or Apple says until they start comparing like equipped servers and THEN compare price/performance


LK

[ribitch]

Originally posted by LegendKiller
Again, what uses are you taliing about? What defines an "average life span"?

According to all finance people I know, the average lifespan of a computer is 3 years, and thats what they use to depreciate them, at the end of those 3 years they consider them scrap, no matter what the computer.

What defines a life span? The speed of the programs being run. What defines what is a good speed? The user of the computer. What affects the speed perception? For one, your own home computer and what yo udo with it, how much money you spend on it. The reverse can be true also. You buy a new PC at work and your home seems slow and you want to get a new one. Since PC's are cheaper you think its ok so you do it. Your 3 year investment now turns into a 1 year resale because of perception and the speed at which PC's evolve.
The same cannot be said of Macs...which have just broken the GHZ barrier...and only that through pipeline lenghtening.



Heck, right now I am on a 5 year old P2-450 with a old ATI rage video card and 256mb of ram. Wow, 5 years. I asked how much these things cost new...1,600 bucks. 320 bucks per year...not bad.



in 1998 (4 summers ago) i boughy a k6-2 300 when they first came out. I may be wrong, but the P2 want that fast 4 years ago, so you didnt have it 5 years ago. Why do I knoiw this? Because I just got out of high school and bought my computer for college. Once again, you shoot yourself in the foot and loose credibility.

As you probably dont know, the 2 CPUS are of different architectures. Apples only at 1 GHz, yet to is as capable as an overrated P4 system. Even AMD now knows that intels speed ratings are bull.

[jase71]

Originally posted by LegendKiller
The major point I was making with the standards was that you can find more or less the EXACT same thing for both computers. You can always find an Adaptec U160 SCSI card for either system, so why not use it to level the playing field, same thing with video cards, a GF4 chipset will perform more or less the same on both systems.

But what good does it do the end user if you build two exactly the same systems that aren't available for purchase?

Why would a potential buyer care what the performance is like with an Adaptec U160 in an Apple server if the U160 isn't an option to buy from Apple?

Would you care how fast your Chevy is with a Ford 5.0 under the hood if you can't BUY a Chevy with a Ford 5.0 under the hood? Of course not... you wanna know how fast your Chevy is with a Chevy engine, in comparison with the Ford with a Ford engine.

Besides, a 1.4Ghz cpu is a 1.4ghz cpu. PC2100 memory is, more or less, PC2100 memory. Motherboards have very little variation in performance.

The performance differences come from two places, the addons, like the controller cards, drives, and nics, and in the case of different platforms, the OS.

If you take away the things that make the boxes different, then what the heck is the value of the comparison? Just so you can find out they bench out about the same? Yippee. I could have told you identical boxes would bench out about the same.

Buyers are trying to choose based on the DIFFERENCES, not on what makes them the same. If they were exactly the same, it wouldn't matter which one they chose.

By trying to eliminate the differences, all you do is get rid of what makes each system unique. You turn an interesting choice into a "flip of the coin".

[ribitch]

Originally posted by LegendKiller

Why would they do this? because they love unfair comparisons with higher-priced systems.


apple wasnt the one who DID this comparison. understand that. get it in your head.

Apple is developing a decent storage solution. Nobody is telling Mike Dell that he cant develop one if they havent already. Since apple isnt using SCSI, they cannot be a viable server contender? I think not. They have a working product, and will soon have another and that is putting pressure on parts of the market. Thats life. Since you have that masters degree, you should know this though.

[LegendKiller]

A working product and a successful product are 2 different things. In order to make their product be successful they need to be able to product parts that the market wants.


Sure, you cannot buy a ford 5.0 mustang engine in a chevy full size sedan. However, taking said ford and comparing it to a more expensive but less powerful 4.0 is NOT a good comparison. Again, 2 different products aimed at 2 different markets but making a comparison to make one look good.

I admire your points, and you have some very good ones, there is no doubt about it. But I know for a fact that until apple can be compared to apple (not the companies) then Mac servers will not be taken seriously by ANY company. These companies (including the one I work for) are not as fickle with their cash, and even less so with their IT systems. Simply saying the price/performance is better than the dell doesn't cut it for them. They need the WHY's and the HOW's. If they see the mac is cheaper and performs better but does it with IDE, the will ask where is SCSI. If Mac doesn't have SCSI raid then, Mac is screwed. BUT, if mac DID have scsi raid, then there goes the price AND performance advantage they have. IT professionals will realize this and any benchmarks that Xinet or Apple uses will be seen as ludicrous and will only diminish both companies credibility.


Its like apples "its a super computer" campaign. Sure, it might have sold more macs to people who thought they were getting a "supercomputer" but for people who knew that it was only the subsets that produced this "speed" it was a big fricking joke. Take that, multiply it by a couple thousand and you will just start to imagine what actual large companies would think of that.

As I keep saying, Apple's products have VERY good merits. They are even very innovative and I would love to see him succeed. HOWEVER, they will not until they can contend with the big boys through REAL benchmarks without using underhanded means of being faster/cheaper.


LK

[LegendKiller]

Ohhh, and a hard drive is a commodity, same with RAm. They do not make a system unique. What makes a computer truly unique is its processor and mobo and the interfaces between the memory and CPU bus. There are several other things. Sure, 4-drive IDE raid makes the Xserve unique, but in a bad way for the IT community.


I hope that apple does well with the Xserve, but they are going to need more than colors and 4-drive IDE "poor mans sportscar" raid to do it.


LK

[DoPeY5007]

Originally posted by LegendKiller
Sure, you cannot buy a ford 5.0 mustang engine in a chevy full size sedan. However, taking said ford and comparing it to a more expensive but less powerful 4.0 is NOT a good comparison. Again, 2 different products aimed at 2 different markets but making a comparison to make one look good.




LK

Unless that ford 5.0 mustang was put next to a 4.0 Camero, then it would be a good compairson, as long as the two prices were about the same

[ribitch]

Originally posted by LegendKiller
Ohhh, and a hard drive is a commodity, same with RAm. They do not make a system unique. What makes a computer truly unique is its processor and mobo and the interfaces between the memory and CPU bus. There are several other things. Sure, 4-drive IDE raid makes the Xserve unique, but in a bad way for the IT community.


I hope that apple does well with the Xserve, but they are going to need more than colors and 4-drive IDE "poor mans sportscar" raid to do it.


LK

complain all you want right now. Sure its IDE. Sure they will have fibre channel. Apple is just touching the tip of the iceburg with this server. It only began shipping less than a week ago. Apple needed to start someplace, and they needed to catch peoples attention. This surely caught your attention. Apple listens to what its users want. If they want RAID in their servers, then it more than likely will happen. With the newness of this server it is too early to tell what they will do.

Macworld coming very soon. We can all expect more from apple then. Will more details on the fibrechannel be given? we dont know. until more benchmarks are released, this one stands as apple beating dell. There is also a benchmarlk of the xserve running apache and giving 60% more connections than that of a IBM server (P4 i think)

[ribitch]

Well, when it comes to TCO comparisons, Apple has that hands down. I don't care to support that at this time (too complicated and long winded for a posting) but I will point out that the price comparisons didn't take in to account the "licence tax." XServe has no client licence expenses (unlimited users) and the OS X Server software is "free" (the software is included in the price of the hardware). Of course we know any form of Windows license is anything but free. I would like to know how many seats were supported by this Dell system.
I don't want to even guess the prices on those Sun systems! It was nice to throw them into this bake off, but those machines are not even close to the cost of the Dell/Apple servers, I would think....


and



For information only... the Sun Fire 4800 Server costs $375,245.00. while Apple Xserve costs $4,500.00. This kind of basic comparison is probably too "basic" compared to reality, but....
Question 1: Considering the 14 pages per minute the Sun Fire 4800 Server is able to deliver, two Xserve could do the job. How much would you have to spend for such a system?
Answer: $ 9,000.00
Question 2: Considering the price of the Sun Fire 4800 Server, how many Apple XServe could you buy?
Answer: 83 !
Comment: probably stupid, but in this situation, you could serve 581 pages per minute...


and



http://news.macunlimited.com/news_123.html
granted, this is just about Apple's prosumer systems and not specifically for the Xserve, but the TCO for the company's systems has historically been found to be much lower than their competitors. I am sure Apple continued this when developing the Xserve.
Already much cheaper is the software. Xserve's preinstalled MacOS X Server can handle unlimited users. I was able to go to Dell's website and configure a system (though I don't know how similar, but I got the price, HD size, and # of proccessors to match... that is all the info we were given).
Now, I don't doubt the validity of Xinet's reporting, but they say that the Dell was "running Windows 2000." They say about the price of the Dell with dual processors: "108 GB of hard drive space carries a price tag of about US$4900."
Good so far
Then they run a test against the Xserve, writing "The Xserve managed about 85 MB/sec of throughput with 10 clients, while the Dell buckled under heavy load and delivered only about 40 MB/sec with the same number of clients."
Now, on the Dell site you have two choices for MS Windows 2000 Server a 5 client license and a 25 user license. To do a test with 10 users, they probably had to purchase at least the 25 user license. That costs $2496 on the Dell Small Business site.
That bumps the price of the system with the absolute MINIMUM configuration that still meets the specs Xinet sets for us (dual processors, 108 GB of space, Windows 2000 Server) to US$6668.00
And I mean minimum... I gave it the slowest processor offered (1.2 GHz). no chassis rail to install (something the Xserve comes with standard), a paltry 256MB of RAM, the cheapest service plan, etc.
Where did Xinet get their prices (I want the discount they got) ? Or did they configure the system with just the 5 user license for price, and then use the 25 user license for testing?
That is where the Xserve costs less immediately!
$4500 for unlimited users VS $4900 for 5 users
$4500 for unlimited users VS $6668 for 25 users


all taken from http://www.newsfactor.com/perl/board/mboard.pl?board=nfntalkback&thread=4860&id=4892&display=1#message_4892 with many other good points

[jase71]

Originally posted by LegendKiller

Sure, you cannot buy a ford 5.0 mustang engine in a chevy full size sedan. However, taking said ford and comparing it to a more expensive but less powerful 4.0 is NOT a good comparison.

Isn't it? If the 5.0 and the 4.0 are close in price, which would you choose?

It's about bang for the buck. Whether it's fair to compare them or not, if I can get a 5.0 for the same basic price as the 4.0, I know who's getting MY money, at least unless there's some other concern that's even more important to me.

It doesn't MATTER if it's fair. If one's a better deal, that's where I'll spend my money.




I admire your points, and you have some very good ones, there is no doubt about it. But I know for a fact that until apple can be compared to apple (not the companies) then Mac servers will not be taken seriously by ANY company. These companies (including the one I work for) are not as fickle with their cash, and even less so with their IT systems. Simply saying the price/performance is better than the dell doesn't cut it for them. They need the WHY's and the HOW's.

Sure. But again, this doesn't apply to the benchmarks. It doesn't invalidate the benchmarks. It just means that benchmarks won't be the primary concern for Corporate buyers.

But then, benchmarks are almost NEVER the primary concern for corporate buyers. So it doesn't matter if the Dell won or lost from that perspective. Nothing would change, even if the Dell was faster.

Cost, service, discounts, support, and kickbacks and freebies are usually much more important in the purchasing decision.

The initial target market for the Apple servers will be small businesses, graphics shops, schools, and places like that. Places where the easier Admin of OSX, the lower cost, and the more "bang for the buck" have a higher value to them.

Target Market.




Its like apples "its a super computer" campaign.

I laughed at that campaign. It didn't sway me a bit.

But apparently, it worked for Apple's target market at the time.
That's what it took to sell more Macs for them.

However, I'm not so sure the Dell Dude is going to help Dell make great inroads in Corporate sales either.




As I keep saying, Apple's products have VERY good merits. They are even very innovative and I would love to see him succeed. HOWEVER, they will not until they can contend with the big boys through REAL benchmarks without using underhanded means of being faster/cheaper.


"Underhanded means of being faster/cheaper."?

God forbid anybody should be faster and cheaper, eh? I'll take my systems slower and more expensive any day! :hmm:

Apple's servers are a great first step for them. Keep in mind, they're new to the game. Dell's been in the business for years. Apple is just getting it's feet wet. They should get better as they gain some experience. They do have some ground to make up, though.

I'm impressed by them, considering they're a first attempt. For jumping into a new market, with no history behind them, they did pretty well.

Am I quaking in my boots about my client switching to OSX servers? No. And despite them being nice servers, I'm sure not going to go out of my way to recommend them to my client. Not when my employer sells PC-based boxes. I don't feel like promoting my own obsolesence.
Nor do I think they're suitable for the needs of my clients yet, anyhow.

But they are nice machines.

[jase71]

Originally posted by LegendKiller
Ohhh, and a hard drive is a commodity, same with RAm. They do not make a system unique. What makes a computer truly unique is its processor and mobo and the interfaces between the memory and CPU bus.

If hard drives are commodities, then why do you review them? Why do you benchmark them?

Is it because there are differences between them? Do they vary in speed and reliability? Or are they all just the same?

Same with RAM. If RAM is just a commodity, why is it I can spend dramatically more for one stick of 512MB DDR than another? Is one stick better, faster, more overclockable than the other? Or are they the same?

What makes a pc truly unique is the combination of it's parts. Not just a narrow subset of them. A killer pc with a dog of a hard drive is a crappy pc. A killer pc with a crappy video card still sucks.

Motherboards seldom vary in performance more than a couple of percent. But video cards, hard drives, and other parts can vary greatly in performance.

[SnowSurfer]

waa waa waa....somebody call the wambulance. ok look the apple server beat the dell. they werent configured the same but they were priced the same. me personnaly i dont care how much i have to pay for a comp. if i get what i need and use then great fine with me. but apple does not offer dells customizabilty on desktops does it? correct me if im wrong but the only geforce4 out for macs is the geforce4 mx. i used to be a loyal apple user but it lost its glamor. sure the ipod is cool. maybe they will get pc software for it but u know what i might even buy it. but the apple desktops and servers (contrary to popular belief (ribitch and others) are not for everyone and not universal like a pc. it just isnt happening yet. you know maybe the site did unfairly compare but some people are stubborn (myself included) and really dont care cause i know what should have been done and who really would of came out on top if it was a full "server war". and ribitch can u tell me of any other companies that sell mac desktops? that is a losing point for them having only one company and no competition (on macs) lets them set the price at whatever the heck they want.

[ribitch]

Originally posted by SnowSurfer
waa waa waa....somebody call the wambulance. ok look the apple server beat the dell. they werent configured the same but they were priced the same. me personnaly i dont care how much i have to pay for a comp. if i get what i need and use then great fine with me. but apple does not offer dells customizabilty on desktops does it? correct me if im wrong but the only geforce4 out for macs is the geforce4 mx. i used to be a loyal apple user but it lost its glamor. sure the ipod is cool. maybe they will get pc software for it but u know what i might even buy it. but the apple desktops and servers (contrary to popular belief (ribitch and others) are not for everyone and not universal like a pc. it just isnt happening yet. you know maybe the site did unfairly compare but some people are stubborn (myself included) and really dont care cause i know what should have been done and who really would of came out on top if it was a full "server war". and ribitch can u tell me of any other companies that sell mac desktops? that is a losing point for them having only one company and no competition (on macs) lets them set the price at whatever the heck they want.

you're wrong
Apple has both the GF4MX & the Ti. Apple introduced the GF4 with NVidia before PC manufacturers introduced it. As far as customizablilty, you can pick nVidia and ATI cards, Your choice of otpical drives, your choice of HD sizes & numbers, your choice of displays, your choice in ram.

Apple produces their own systems. You used to have the ability to make clones, but they lost their stability, so apple decided no longer to support clones. You can also buy apple systems from many vendors. Can you go to circuit city and buy dell? No, but you can get an apple. Can you go to the gateway stores to get a dell? Nope, sorry.

A computer manufacturer usually only sells their OWN systems. Thats the way business works. Apple decided stability and reliabilty was most important to its OS users, so it no longer supprts clones.

OS X has windows/mac/unix file and print serving set up right out of the box, plus there is no server tax like other vendors have. OS X and XServe are not made to be a windows authenication server. So yes its not for eveybody. However, if a company is looking for an affordable, reliable, fast server for filesharing/print serving/database serving/webserving, or kerbos authenication and they have yet to get into the rack mounted arena, then apple is a decent contender for their business.

[SnowSurfer]

Originally posted by ribitch


you're wrong
Apple has both the GF4MX & the Ti. Apple introduced the GF4 with NVidia before PC manufacturers introduced it. As far as customizablilty, you can pick nVidia and ATI cards, Your choice of otpical drives, your choice of HD sizes & numbers, your choice of displays, your choice in ram.

Apple produces their own systems. You used to have the ability to make clones, but they lost their stability, so apple decided no longer to support clones. You can also buy apple systems from many vendors. Can you go to circuit city and buy dell? No, but you can get an apple. Can you go to the gateway stores to get a dell? Nope, sorry.

A computer manufacturer usually only sells their OWN systems. Thats the way business works. Apple decided stability and reliabilty was most important to its OS users, so it no longer supprts clones.

OS X has windows/mac/unix file and print serving set up right out of the box, plus there is no server tax like other vendors have. OS X and XServe are not made to be a windows authenication server. So yes its not for eveybody. However, if a company is looking for an affordable, reliable, fast server for filesharing/print serving/database serving/webserving, or kerbos authenication and they have yet to get into the rack mounted arena, then apple is a decent contender for their business.

can u pick the soundcard? the certain brand of cd-rw/dvd drive? the number of fans?

oh wow i can buy apple systems from many vendors...yes circuit city may have them but they dont have customibility do they? and why would i go to the GATEWAY store if i wanted a dell or a apple. u totally missed the point i was making. the point i was making was the windows os is on very many vendors systems. apple os is on APPLE systems. (for example, windows is on Dell, Gateway, Alienware, Hp...ect. apple is on apples and only apples.)

yes computer manufactuers sell their own systems but do they sell a os. i dont see the dell os or gateway os anywhere.

good u finnaly said it. it is not meant to be a windows authetications server so why would someone running windows systems that is new the rack mounting area and wants a server buy this? they wouldnt. we will see their selling numbers if they even release them.

[ribitch]

Hate to break it to you, but windows isnt the only platform used in the business world. Many companies refuse to use windows due to its poor security, poor stability, and prone to virus's.

People also buy servers for things other than authenication. You only need 1 system to authenicate. If you have several people authenicating at one time, you may split it to 2 systems with load balancing. Companies often have dedicated systems for other tasks. Thats where the xServe comes in.

What if they DON'T use windows? Xserve can authenicate using kerbos, and this allows platforms other than mac and windows authenicate.

Sure, you cant customize an apple at compusa (they do offer ram upgrades and HD upgrades), but can you customize your alienware system from bestbuy? Not really. They often send the system back to teh manufacturer to get the new items in. If you want a customized system, you get it from the manufacturer. You can do this with apple.com and apple stire, or many other apple e-tailers.

As far as the need for several fans, macs dont have heat problems like a PC does. Their cases are designed with airflow in mind. Cables are routed through the case in ways that would make a PC overclocker moist. There was a company in maximum PC that has some origami guy doing the twisting and routing of the cables in their systems. They made it look like it was an industry first. Apple has had this for years.




Originally posted by SnowSurfer


can u pick the soundcard? the certain brand of cd-rw/dvd drive? the number of fans?

oh wow i can buy apple systems from many vendors...yes circuit city may have them but they dont have customibility do they? and why would i go to the GATEWAY store if i wanted a dell or a apple. u totally missed the point i was making. the point i was making was the windows os is on very many vendors systems. apple os is on APPLE systems. (for example, windows is on Dell, Gateway, Alienware, Hp...ect. apple is on apples and only apples.)

yes computer manufactuers sell their own systems but do they sell a os. i dont see the dell os or gateway os anywhere.

good u finnaly said it. it is not meant to be a windows authetications server so why would someone running windows systems that is new the rack mounting area and wants a server buy this? they wouldnt. we will see their selling numbers if they even release them.

[ribitch]

the original point of this thread was to show that the dell had came in second to the apple. It wasnt about saying switch to apple and you're life will be better or anything of that nature. Then people who didnt read the article or report that xinet did started saying bitching saying it isnt a valid comparison and things of that nature, then finally it was can apple do thi sor that.

As in most cases, the PC users state something and usually are totally incorrect (just on the apple facts, not all facts), and I correct them.

Sure, there are fewer choices for video cards direct from apple, or audio choices. But the truth is apples audio is better than that of a sb live, and if it isnt what you need, you buy something from a company like this http://www.digidesign.com/

If the options for a video card dont do what you want it to do, then once again, you can order something like this http://www.matrox.com/video/products/rtmac/home.cfm

There are many choices for apple hardware upgrades that range from entry level consumer to highend professional systems. Apple doesnt use low quality hardware, so they elliminate the need to "upgrade" the manufacturer of your optical drive and things of that nature. Instead of giving their customers a low quality sound system, apple R&D have developed a system that most musicians dont even upgrade when mixing/editting their music.

The basis of the "Apple systems cant be upgraded/customized" is totally untrue. The reason that apple doesnt offer as many choices on customizing is their hardware is not low end, and if it is too lowend for your wants/needs, then you will often be seeking a specific solution for your needs.

I have not told anyone in this thread that they would be better off on a mac. I have given reasons why an Xserve could be used, links supporting some of my statements, and the truth to misconceptions about apple systems. In my eyes, a gaming system is a console. A computer also does a fine job at gaming, but the console is still superior since it doesnt need the frequent upgrades required to play the same games as on the PC. If you prefer to pimp out a PC to play your games, more power to you. All I ask is that you dont slam a mac/PC/whatever until you actually give it a shot and try using it.

[Mpowered]

Originally posted by LegendKiller
A working product and a successful product are 2 different things. In order to make their product be successful they need to be able to product parts that the market wants. I hate this argument sooooooo much. Saying if a mac is a better product, why do most people buy windows is similar to saying if Mercedes are better, why do so many more people buy Caddilac's? sure a Benz costs more, but a good portion of Caddie owners can afford the Benz. So why do people do go for the Caddie? Could it be that its a sturdier car? Nope, Benzes are definately sturdier. Could it be that they last longer? I personally have never seen a 200,000 mile Caddie that was in as nice condition as a benz with the same milage. Maybe its the resale value. Wrong again, Caddilacs lose half thier value in a few years, and Mercedes Benz's have some of the best resalw value in the luxury market or any market except super luxury like Rolls Royce or exotics like Ferrari. So waht to you think? Are Caddilacs better than Mercedes Benz's? You must, Caddies sell more and are more successful

[Mpowered]

Originally posted by DarkFury
It must have something to do with "price" and the cost of ownership huh?

That tends to happen with most "big ticket items"... Sure there is the "top of the line", but when the middle will do at much lower costs, then it gets the lion's share of the buying public.

Some folks just like the "status" that is associated with a name like "Mercedes" and some folks just wanna be able to get from point A to point B reliably and at a reasonable cost. :D

You're right DF, it wasnt a great example. My point was that just because most people buy windows based PC's, doesn't meant that its a better product and that argument is pointless and lame. Anyone got a better example feel free to post it.

[Ladogaboy]

Originally posted by ribitch
I am currently looking for the report on average number of applications used. This isnt talking about apps like minesweeper or AIM, its talking about things like lotus, office, adobe apps, etc.

Just out of curiosity, how can you pass judgements on what kind of apps are being used? If John Doe goes out and buys a computer for $X, he is probably planning on doing certain things with it. If all he does is IM and email, then that is his perogative, but it is also his reason for buying the computer.

I agree, on the big scale of things, solitaire probably isn't as important as Final Cut Pro, or whatever, but what we are ultimately arguing is whether the computer in question can perform the functions that are required of it. By eliminating games and entertainment from the cadre of programs being considered, you are giving Mac an advantage over PCs by eliminating one of the main reasons that PCs are more versatile than Macs. Sure, Final Cut Pro might be better than Adobe Premiere, but when I get done editing a video and want to play a game, which system am I going to be able to plop a Recent game in a know it is going to work?

Since everyone here seems to like car analogies, that is like comparing two luxury cars but leaving the sound system out of the comparison because 'entertainment' isn't as important as the rest of the features. While that might be true, it avoids the point that many people buy cars for more than just getting from point A to point B.

Anyway, getting back to the main point of this whole thread, I still agree with LK. The review was very skimpy in terms of letting us know exactly how the hardware was configured. Sure, it could be argued that they are precofigured systems, but what other options did they have? Jase is right, too, though... you can't possibly completely level the playing field, nor would you want to in this case. I just prefer reviews that are MORE open than this one.

And don't get me wrong about my allegiances, either... I firmly believe that Macs are the best pre-built systems that one can buy, but on the other hand, I don't buy pre-built systems. I like PCs based on their versatility, period. I have considered buying a Mac at various times, believe me, but there are still too many hurtles for me to actually make the plunge. Especially not when I have the option of dumping Gates and still retaining the versatility of hardware and ease of configurations that PCs provide.

[ribitch]

Originally posted by Ladogaboy


Just out of curiosity, how can you pass judgements on what kind of apps are being used? If John Doe goes out and buys a computer for $X, he is probably planning on doing certain things with it. If all he does is IM and email, then that is his perogative, but it is also his reason for buying the computer.

I agree, on the big scale of things, solitaire probably isn't as important as Final Cut Pro, or whatever, but what we are ultimately arguing is whether the computer in question can perform the functions that are required of it. By eliminating games and entertainment from the cadre of programs being considered, you are giving Mac an advantage over PCs by eliminating one of the main reasons that PCs are more versatile than Macs. Sure, Final Cut Pro might be better than Adobe Premiere, but when I get done editing a video and want to play a game, which system am I going to be able to plop a Recent game in a know it is going to work?

Since everyone here seems to like car analogies, that is like comparing two luxury cars but leaving the sound system out of the comparison because 'entertainment' isn't as important as the rest of the features. While that might be true, it avoids the point that many people buy cars for more than just getting from point A to point B.

Anyway, getting back to the main point of this whole thread, I still agree with LK. The review was very skimpy in terms of letting us know exactly how the hardware was configured. Sure, it could be argued that they are precofigured systems, but what other options did they have? Jase is right, too, though... you can't possibly completely level the playing field, nor would you want to in this case. I just prefer reviews that are MORE open than this one.

And don't get me wrong about my allegiances, either... I firmly believe that Macs are the best pre-built systems that one can buy, but on the other hand, I don't buy pre-built systems. I like PCs based on their versatility, period. I have considered buying a Mac at various times, believe me, but there are still too many hurtles for me to actually make the plunge. Especially not when I have the option of dumping Gates and still retaining the versatility of hardware and ease of configurations that PCs provide.

some rather good points. I was even shocked that somebody on this board knew what FCP was. Very well stated. As far as games go, developes seem to be getting more mac friendly with releases now that OS X is here. Maybe this will also help more games be ported to Linux. Only time will tell

[Ladogaboy]

Originally posted by ribitch

As far as games go, developes seem to be getting more mac friendly with releases now that OS X is here. Maybe this will also help more games be ported to Linux. Only time will tell

Yeah, hopefully, but I've been moving away from games myself lately... Too much money, and too time consuming.

I've actually been talking with jase about Linux, and I've started to meddle with it a bit. So far, I'm very impressed.

And, like I said earlier, don't get me wrong, I like Macs, though I am not likely to buy one. Sometimes, I'm a harsher critic of those things I actually support, just because I expect them to hold up to the criticism. It also pisses me off when a product I would like to support doesn't market itself as well as it could. For instance, I am kind of ticked at some of those new Mac commercials that are talking about Windows short-comings... some of the examples, i.e. BSoD, that they give about Windows' failings don't apply to the newer OSs.

From what I've seen, if Mac wants to build a bigger market share, they are going to have to convince the people who actually know something about computers first. I'm kind of using AMD as an example for this... with the release of the K6-2s and 3s and subsequently the Athlons, a lot of informed PC users realized that AMD was a true contender against Intel. Later, that knowledge slowly diffused itself into the rest of the population until even the people who know almost nothing about computers were lauding AMD and asking for them in stores. If Mac really wants to start taking back a chunk of the computer market, they need to convince the people who actually know something first... IMHO anyway...

[jase71]

Originally posted by Ladogaboy

If Mac really wants to start taking back a chunk of the computer market, they need to convince the people who actually know something first... IMHO anyway...

The other day I saw the first Mac commercial that ever really impressed me. It wasn't flashy Imacs spinning in circles in different colors. It wasn't Ghandi telling me I should use a Mac.

It was a bearded programmer geek, a Unix-head, telling me how easy it was to use his OSX machine. He didn't spin, he didn't play funky music. He didn't tell me how to think. It was actually pretty persuasive.

Not that I'm in the market for a Mac. I like Linux enough that I'm not willing to buy Apple's hardware just to run OSX. I like OSX a lot. But I'm not eager to give up PC hardware... and I won't as long as I have Linux. Now, if OSX ran on the x86 platform...

But Ladoga is right. Apple can gain some market share by running funky commercials, building their units in pretty colors, and attracting the hippy/college/artsy crowd. But to gain serious market share, they need to grab mindshare with the tech savvy people.

When Mom and Dad decide they need a computer, they don't buy based on tv ads. They ask their techie son or daughter what to buy. Or a guy at work. And the recommendations follow the market share. 9 times out of 10, the person recommends a pc. So that's what Mom and Dad buy. And that guy probably makes recommendations a couple of times a month. That adds up to a lot of computers bought, solely based on the recommendations of a small group of people.

For years, Apple has needed TWO things before they could grow. The first is the support of techie people. The second was a better OS.

With OSX, they now have the better OS. If you look purely at the OS, it might be the best OS on the market right now, even including my beloved Linux. Even if it's not the best, it's more than good enough to compete with Linux and Windows.

What Apple needs now is mindshare with the geeks. The people who influence other people. That's going to be tough to come by, because it's the market Apple has largely ignored for the last 10 years. They went after the creative market, and ignored the geeks and hobbyists, the techies and the hackers.

Whether or not they can regain that techie support might determine whether Apple will ever be a force again. The recent commercial was a good first step. It'll be interesting to see if it's just one commercial, or if it's part of a broader plan to go after that "geek" support.

[SnowSurfer]

yeah apples dont really seem to appeal to the techie/hacker people. but i dont thinks that is what they had in mind when they made them

[ribitch]

Apple & the geek crowd

some great points again. I think apple is starting to really address this now as well. The first true "geek" ad was a print ad talking about how O SX was unix and things of that nature. It was an ad that caught some peoples attention, but most people just keep flipping past the ads to get to the magazine content.

Then came the programmer switch commercial. Once again another "geek" ad, but this time it caught more people attention. Most people, like myself will usually not flip through the commercials until the break is over. This way seems much more effective than a magazine.

OS X is close to a geeks dream come true. Fast processors (MHz numbers dont justify them), stable customizable OS, stable OS, developer tools (including IDE) free from the manufacturer, developer incentives, and much more. The big hurdle apple is facing once again is price and previous opinions on the product.

The prices have fallen over the past few years. The ibook is very well priced for its features. The powerbook is still out of most peoples price ranges (i wish it would come down), the powermac can be affordable if you get a CRT, and the iMac is amazingly affordable based on the Superdrive version. They still can improve pricing, which I see will slowly happening.

The main hurdle with the geek crowd is opinions like those expressed by many here. Many "geek" types are members of this board. Many have slowly changed their opinions, but many will not budge on their opinions.

Ask almost any geek though what is considered the ultimate laptop, and many will respond the powerbook. What is keeping them from buying it? Price and old opinions on apples in general. Apple was always known for educational software, not MS office, or games like black and white, max payne, sims, the warcraft's, the diablo's, giants, alice, etc. What geeks dont realize is these products are all available for the mac.

Apple know needs to make some "switch" ads featuring the software the geek types will be using. Not all the hacker tools, but the major apps, as well as major games.

[ribitch]

heres another article that disputes the "apple messes with the drivers/ram/etc when doing a photoshop bench"

link (http://www.gannettonline.com/e/eteam/mac_reality_check.html)

the gaming aspect, the PC did get 194 fpd compared to 11fps on the mac, but the PC has the GF4 Ti while they used the GF2MX on the mac (dont ask me why it didnt have the GF4Ti in it)

were the systems close, yes. But once again, its use what suits your needs.

[Cantacuzene]

Originally posted by ribitch

Ask almost any geek though what is considered the ultimate laptop, and many will respond the powerbook.


Heh, close, but no quite... The best laptop (http://www.alienware.com/main/system_pages/area51-m.asp)

[ribitch]

Originally posted by Cantacuzene


Heh, close, but no quite... The best laptop (http://www.alienware.com/main/system_pages/area51-m.asp)

impressive, but still not the ultimate in my eyes. Sure its upgradeable, but what about bettery life, weight, size, height generation, and things of that nature.

A laptop is designed with portability in mind. If the thing weighs 10 pounds, has 2 hour battery life, and takes up tons of space, then it isnt serving the original purpose and the price makes a desktop a more vaible option.

based on the pictures it looks really thick (look at the USB ports on the side view and the parallel ports) This is starting to remind me of that laptop with the external batteries.

This wins for the closest system to a desktop, but not the Ultimate laptop

[ribitch]

upon further review, the system has the following specs:

Dimensions (H) x (W) x (D) : 1.7"x13"x11.4"
Weight: 9.6lbs. w/ Battery

battery gives 1 hour of usage, a second could be used to get an additional hour of usage.

powerbook:
Size and weight
Height: 1.0 inch (2.6 cm)
Width: 13.4 inches (34.1 cm)
Depth: 9.5 inches (24.1 cm)
Weight: 5.4 pounds (2.45 kg) with battery installed
battery life of 5 hours

far from ultimate laptop. Almost double the weight, 70% thicker, deeper and 0.4 inches narrower than a powerbook. The second battery would also add much more weight to it and it would still get less than half the battery life.

[LegendKiller]

The main disadvantage of the Alienware/battery life is that the Alienware/Sagar brands use desktop P4's as their CPU's which consume massive amounts of juice. A much more reasonable alternative is the mobile p4's, they may be slower but they do not kill battery life.

My Dell 8100, 1ghz p3-m, 320mb ram, 20gig, and IS upgradible (GF4 440go in it), and has a 15" UltraXGA (1600x1200 res) screen weighs around 6.5lbs with a battery life of 2:40min, a 2nd one can be added for 5 hours which would bring the weight up to 7lbs (removing the floppy also). I can lengthen that even more by turning the backlight down further, having the HDD park sooner, turning fans off (I can do that manually), and tuning down the GF4go.

Sure, the powerbook may be lighter/smaller, but it doesn't have NEARLY as many features and advantages. I am able to do everything on my laptop as I do on my desktop (games(at very good frame rates), graphics, and word processing), and I can upgrade it just like a desktop. Taking into account its massive and VERY high quaility screen, and its very low price, it is about the best laptop a person can buy, and the 8200 is the same way.

My laptop cost me 1200 bucks brand new. The base powerbook is 2500 bucks, which is rediculous. Dont even start about the Radeon (wait for my notebook article to hit comparing the GF4go to the radeon). And dont even start about Gigabit networking, the average comp geek WONT have a switch to handle that. I also have firewire, internal 56k/10/100, and I have my own wireless. Even if the 667mhz powerbook was quiv to the P3-1ghz (which it isn't), that would still mean I beat it in many other performance/price points.

[ribitch]

Originally posted by LegendKiller
Even if the 667mhz powerbook was quiv to the P3-1ghz (which it isn't), that would still mean I beat it in many other performance/price points.

huh? I am confused on what you meant with this. the 667 is by far faster than a 1GHz P3.

as far features go, is the firewire integrated? or is it taking up a PCMCIA slot? What about the wireless?

What about an optical drive with dual batteries?

5hours on dual batteries. You can also purchase a second battery for a powerbook and get 10hr batterylife.

What about dimensions?

How is it upgradable? Can you replace the CPU?

I have yet to see a laptop that cant have its harddrive replaced or more ram added.

True, most geeks wont use gigabit yet, but in graphic design, it is very beneficial and often used. Same with video production.

Sure, the ATI Radeon isnt a gaming chipset to most people, but the powerbook ISNT a gaming system.

[LegendKiller]

The P3-M isn't your run of the mill P3, it beats out a P3 by about 10-20% alone due to its smaller die, larger cache, and more optimizations, I can assure you that it would whipe the floor with the powerbooks CPU.

Firewire is integrated, and you can get integrated wireless installed, along with 10/100 and 56k. Again, gigabit isn't that big right now, only larger infrastructures have it, and even then it wouldn't make much of a difference on a mobile since you ARE mobile. Thats not that big of a diff.

You can have a dvd/cdrw combination with dual batteries.

Sure, you can cram 10hours from a dual battery, I was just making a direct comparison to get them both to 5 hours.

What do dimensions matter? I can tell you it fits fine on an airplane tray, on a desk it doesn't matter much.

Height: 44.5 mm (1.75-inch)

Width: 331 mm (13.03-inch)

Depth: 276 mm (10.8-inch)

It isn't that much larger.

If I can get my 8100 GF4go/graphics comparison put up this week you will see how upgradible this sucker is. You can ramp up the CPU to 1.4ghz and you can change the video card from a Radeon M4-7500 or the GF2g0-gf4go since it has an AGP slot. The 8200 also has this feature with both the CPU and GPU. The CPU slot can even take desktop procs.




Originally posted by ribitch


huh? I am confused on what you meant with this. the 667 is by far faster than a 1GHz P3.

as far features go, is the firewire integrated? or is it taking up a PCMCIA slot? What about the wireless?

What about an optical drive with dual batteries?

5hours on dual batteries. You can also purchase a second battery for a powerbook and get 10hr batterylife.

What about dimensions?

How is it upgradable? Can you replace the CPU?

I have yet to see a laptop that cant have its harddrive replaced or more ram added.

True, most geeks wont use gigabit yet, but in graphic design, it is very beneficial and often used. Same with video production.

Sure, the ATI Radeon isnt a gaming chipset to most people, but the powerbook ISNT a gaming system.

[ribitch]

1.75 inches thick!!! thats 75% thicker than a powerbook. Briefcase/backpack space is a premium. Having a laptop take up nearly double the room is a big disdvantage.

if you can upgrade the videocard by yourself, then I do give it credit for having an AGP slot. If its a factory option only then its not as upgradable as you say it is. How readily available are the vidoe card upgrades? How affordable are they?

as far as upgradability goes, you are all about the upgradability, yet you blow off gigabit. That is something that wont need to be upgraded down the road on the powerbook, while the dell will need it.

With 2 batteries in your dell, you have a system that is 2 lbs heavier than the apple, and almost twice as thick, which means more of a burden when traveling with it.

[ribitch]

i forgot to comment on the integrated wireless on the dell. I remember when micheal dell introduced it. He boldly claimed that they were the first manufacturer to offer it. WRONG mikey, apple had that first.

Just wanted to mention that

[LegendKiller]

That really isn't that much thicker, my bag's hold it just fine.

No, they didn't offer it as a factory replacement, but you can attain them pretty easily and the GF4go module is only 150 bucks.

Upgrading a processor/gpu is MUCH different from upgrading networking thats why you have PCMCIA cards, to handle that kind of thing. A PCMCIA video card or CPU is not possible.

Again, another 2lb's isn't that big of a deal (at least not for me...but then again I am 6'2" and 240lbs with a fair amount of muscle). And with a good backpack/shoulderstrap that is even less. I mean, people bitch about lugging 6 lbs, there are soldiers or even ordinary people who take 50-70lb backpacks on hikes.

Thats half of the problem with today's society, gotta be afraid of working the muscles...


LK






Originally posted by ribitch
1.75 inches thick!!! thats 75% thicker than a powerbook. Briefcase/backpack space is a premium. Having a laptop take up nearly double the room is a big disdvantage.

if you can upgrade the videocard by yourself, then I do give it credit for having an AGP slot. If its a factory option only then its not as upgradable as you say it is. How readily available are the vidoe card upgrades? How affordable are they?

as far as upgradability goes, you are all about the upgradability, yet you blow off gigabit. That is something that wont need to be upgraded down the road on the powerbook, while the dell will need it.

With 2 batteries in your dell, you have a system that is 2 lbs heavier than the apple, and almost twice as thick, which means more of a burden when traveling with it.

[LegendKiller]

Originally posted by ribitch
i forgot to comment on the integrated wireless on the dell. I remember when micheal dell introduced it. He boldly claimed that they were the first manufacturer to offer it. WRONG mikey, apple had that first.

Just wanted to mention that

Kinda like how jobs boasted about being the first to have GeForce3 cards installed but didn't start shipping computers with them until WEEKS after dell?


Funny how people say what isn't true...or even how the G4 is a "supercomputer"...BWAHAHAHHAHA, yeah, maybe compared to a 20 year old cray. It doesn't even come close to touching even a first gen Itanium.

LK

[ribitch]

Originally posted by LegendKiller


Kinda like how jobs boasted about being the first to have GeForce3 cards installed but didn't start shipping computers with them until WEEKS after dell?


Funny how people say what isn't true...or even how the G4 is a "supercomputer"...BWAHAHAHHAHA, yeah, maybe compared to a 20 year old cray. It doesn't even come close to touching even a first gen Itanium.

LK

apple was the first to intro the new geforces (GF4). problems at nvidia delayed the shipment

link (http://www.hardwareanalysis.com/content/article/1473/)

supercomputer by definition is the ability to perform 1 gigaflop. Apple was the first consumer desktop system to achieve that. The definition was originally definded by the government not apple. By releasing a system that exceeded that 1 gigaflop barrier, apple had essentially released a supercomputer. A current apple desktop is capable of 15 gigaflops.

The itanium 2 is supposed to be able to achieve up to 13 gigaflops. The G4 already has that beat.

Why doesnt intel or AMD base speeds on gigaflop capability since it can be a real world performance benchmark?

[LegendKiller]

LOL, basing gigaflop performance on the Altavec performance is hardly a measurement of TRUE desktop performance. Its like saying "Oh yeah, my Toyota can get up to 180mph for 5 seconds, thats just as good as a Ferarri". Sure, it can get to 180, but it can't stay there for any period of time.

The throughput of ONE certain portion of a processor (SSE2/3dnow, Atlivec) hardly makes a supercomputer. I have yet to see benchmarks to prove that any of the current gen's are equal to a "supercomputer". Even funnier is the fact that 1gflop is not even that big of a deal, a "supercomputer" of 20 years ago could do that. Another marketing gimmick that real people knew was stupid while stupid people thought it was real.

Sure the Powerbook is an awesome computer with many great features, but I think that its not even a contender in the real world until they drop the price by 800 bucks.

LK

[ribitch]

Originally posted by LegendKiller
LOL, basing gigaflop performance on the Altavec performance is hardly a measurement of TRUE desktop performance. Its like saying "Oh yeah, my Toyota can get up to 180mph for 5 seconds, thats just as good as a Ferarri". Sure, it can get to 180, but it can't stay there for any period of time.

The throughput of ONE certain portion of a processor (SSE2/3dnow, Atlivec) hardly makes a supercomputer. I have yet to see benchmarks to prove that any of the current gen's are equal to a "supercomputer". Even funnier is the fact that 1gflop is not even that big of a deal, a "supercomputer" of 20 years ago could do that. Another marketing gimmick that real people knew was stupid while stupid people thought it was real.

Sure the Powerbook is an awesome computer with many great features, but I think that its not even a contender in the real world until they drop the price by 800 bucks.

LK

the gigaflop may not be the best judgement of speed, but intel is revertingf to using it to try to prove MHz isnt everything anymore. The Itaniums 2's run slower than a P4, yet are a faster overall chip (I have never used one, so i dont know for sure) So intel is using the gigaflop as a speed indicator. is there a better way to compare systems since MHz isnt a usable factor anymore? Maybe.

As far as a real world contender goes, I will leave you with your opinion. You liek dell, so be it.

[LegendKiller]

I could be wrong, but the gigaflop measurement that Apple was quoting was only while using the throughput from the Altivec processor, which was a SIMD processor, much like MMX, MMX2, SSE, SSE2, and 3dnow. They are only streaming certain instructions and not the whole processor.

Furthermore, its not that I like dell (for my laptop at least...I build my own desktops)more than any other computer brand, I like what gives me the most bang for the buck. Thats the way finance works. You find what is the best hardware, you figure out how long it will last, and the return on investment, and you make a decision.

When a Mac desktop or laptop can be as upgradable as a Dell, as cheap as a dell, and do everything I need out of my computer, then I would switch if it provided a better deal. However, they cannot play all of the games I play, they cannot be upgraded the way a dell/PC can, and they carry an unacceptable price premium for what you get/dont get.

Thats the way the rest of the world sees it also. Combine that with the elitist/fanatical/fringe element image that Apple carries and you have just niched yourself into about 5-10% of the worlds population, right where apple is right now. They are totally missing joe-sixpack. Thus, they will never get big enough sales to push their innovation into mainstream and will never hit the even bigger bux of the server market.

Its actually quite sad, while their processors and architecture may be superior it is hindered by the "fanboy" mentality of its users and CEO. If only the mac could get passed these roadblocks and penetrade out of the corner it is in, then everybody would want one.

Until then, its a second rate machine.

LK

[ribitch]

once again you bring us back to the cost basis of the computer. You say most bang for the buck and things of that nature, but you never fully work out the big picture. Things like energy consumption, hardware longetivity, and reslae value are all overlooked. Compare this once again to the cost of a car.


lets compare a 800 MHz imac
energy consuption - gas cost. a system that eats up say 260 watts under max continuous use costs more to perate than that of a system using 130 watts under continuous usage. lets use the system 10 hours a day. energy costs at 7cents.

130*10 =1.3Kw per iMac per day = $0.091
260*10 = 2.6kW per PC per day = $0.182

hardware longetivity - You usually buy a car that will withstand time. You dont buy a car that dies after 2 years (like the dell upgrade cycle noted at earlier link or a car that will last 3.9 years. If the car that lasted almost double the time of the other was slightly more (imac remember. Prices are about teh same between imac and Dell with 15"flatpanel) What would be the more econmical choice?

now you have had your use from the car(system) and its time to sell it. (midpoint between 2 years and 3.9 years is about 3 years) So after 3 years you decide to get something new. What would you get more money for? The mac actually retains the resale better than the Dell.

Now in the long run, what is the better value. Deny it all you want, but its the imac. As far as costs go, price out a DVD burner equiped Dell with a 15inch lcd display and optical mouse. Also include an office style app (apple include Apple Works free of charge), a mp3 player (iTunes), DVD authoring software (iDVD), and a photo software (iPhoto), and a decent set of speakers (apple pro speakers) see what the price comes to. keep it at a simple 256 MB ram, 60 GB harddrive, and 32 MB video card. Also OSX is a pro level OS, so include XP pro, not home

so the specs should include:
fastest available entry level processor
256 MB Ram
60 GB HD (may go lower if not offered to save money)
DVD Burner
Speakers
15" LCD
photo suite
movie authoring package
Office style application
Mac OS X/Win XP Pro
MP3 Software player

Not rebates etc. on pricing. No educational discounts.

the mac comes in at 1900 dollars. Wheres the dell. (when I did it it came in over 2100, but i will let you post it.)

P.S. these specs will probably be bumped up a notch on July 17th for macworld)

[Apex]

More benchies:

BareFeats "BARE Facts on Macintosh speed FACTS" (http://www.barefeats.com/pentium4.html)

[LegendKiller]

LOL,I hate to tell you dude, thats the LAST thing on a companies mind when they go to order a computer. What they want to know in costs is maintenance, performance, and features. They dont give a flying dog crap how many watts it eats.

Another thing I find funny is you say the mac holds it resale value better than a mac. THis is hilarious because you have almost nowhere near any business experience. Businesses just dont say "Well, I think we have had enough of this computer, I guess we will go sell it to bob down the street". Companies do not sell computers, they either hand them down to lower people, or they SCRAP them. In that case, no matter WHAT computer you buy, it has ZERO resale value. As far as an accounting point of view, once the computer has depreciated to 0 value is junk. They do not say "look, I think we can make computer X last longer than computer Y". Standard accounting practices in most companies say a computer will last 3 or so years. After that then the computer is too slow to run an application at needed speeds. This goes for ANY computer. Some companies, like the one I work at, extends this past 5 years (I used a P2-450, so take that buddy).


Actually, I was able to get a Dell Dimension 4500s for 1600, Celeron 1.7ghz, 256ddr, 40gig (they dont offer 60's, and most businesses would find it useless anyway), 15" LCD, dvdRW. Wow, what do you know, 300 bucks less and you get dells reknown business association along with a good warrenty and a computer that will work seamlessly with the rest of the world. At my work, they sometimes use XP home, and I would hardly call OSX a "orofessional" os, most companies laugh when you say that. Even with XP pro thats only 99 bucks more. Most companies will use their own OS licenses which are bulk cheap


You say you shouldn't use rebates and such...why? We have to compare price points dont we?


Also, as Apex pointed out...the apple got smacked by the PC...

[ribitch]

a celeron isnt comparable to the G4. Therefore the P4 systems were used. Win XP Home cannot be added to a domain, therefore a centralized authenication server cannot be used, therefore isnt a business class operating system. OS X and XP pro can all have a centrallized authenication server, therefore are an option in a business.

You talk about mac prices being to high for you to purchase a system. So I give you facts to prove they are cheaper than a PC in the long run. So you say well a business never looks into thinsg like that. Your business may not, but when a company has 50 systems, they see the long term benefit more than a short term. Thtas why some companies are ditching CRT's in favor of LCD's. The energy used is much less, offestting the extra expense.

Your "insider" knowledge of a working business is pretty weak. Apparently your companies IT staff isnt really looking at overall costs of a system. Many large companies require this sort of data, apparently yours doesnt.

As far as the benches go, those benches are a few months old. I have seen them before. The mac was on top of the photoshop tests. The 3D tests (Bryce & quake) proved the non mac systems were best for 3D tests. Nothing new. A mac isnt a high end 3D system.



Originally posted December 31st, 2001
Updated February 27th with Athlon results from "remote mad scientists"
Updated March 4th with G4/933 results

Prices of the PC's were comparable to the prices of the mac 7 months ago. Powermacs are expected to be refreshed on the 17th. DDR or a G5 is expected if any changes are made, and the playing field will be leveled once again.

PC's are updated constantly. Macs on the other hand are updated an almost fixed length interval. (macworlds)

Also note that photoshop 6 was used on the photoshop tests. They say OS X was used, photoshop 6 is not an OS X app. It is using classic mode (an emulator of OS 9). There is a large speed gain from using a native OS X app, such as Photoshop 7. I am curious on how the photoshop tests will fare against the fastest PC if photoshop 7 was used.

[Apex]

Photoshop 6 vs photoshop 7 speed tests:

Faster, but not greatly so, according to Bare Feats. (http://www.barefeats.com/foto7.html)

[Cantacuzene]

I'm sure I speak for a lot of people when I say 3d preformance is more important than Photoshop anyway. I have a PC because it plays games, which are 3d intensive. I use photoshop once a week for menial tasks.

I don't think cpu benchmarks tell the story at all. No computer shopper says, "Well, the PC can play all these games and use all these programs, but the Mac can apply a PS7 filter .0002 seconds faster, so I'll buy the Mac!"

I'll accept the Mac may save you a few cents in electricity or something liek that, but who really cares. Everyone has a list of priorities in what they want their computer to be and I'm bettign Photoshop filters and power consumption are not high on people's list.

[LegendKiller]

Companies aren't going flat panels for JUST that reason dude, trust me. They are doing it for many different reasons, space, weight, eyestrain, radiation, and power also.

I analyze the financial side of IT decisions, I am doing that for a major IT company, and I do it quite well. I can tell you that power consumption isn't that big of a decision. If I were to rate performance, reliability, longevity, features, price, and power consumption/savings on a scale of 1-7, power consumption either would be a 0, or last, thats just a fact, its a null argument.

Furthermore, a 800mhz Imac would actually be equal to a low end P4, which was 50 bucks more, still 150 cheaper. Besides, company's dont look at low-end systems, they look at the high-end to get more longevity out of it. Why buy a system thats already slow? It doesn't make financial sense.

The one I configured also had a LCD, and was only 1700 with XPpro, as I stated.

LOL, Photoshop, I thought the Mac always won in photoshop, then you go and make some statements to say why it ISN'T in ONE CERTAIN version. Dude, and how can you say the Mac was on top when the AMD 1.4 beat it in every test?

Sure, I cannot convince you that you are being a fanboy and not a very logical one. But then again, I just get more ammunition for burying Macs from any IT decision made within my company. So far you have eliminated Xserves, Powerbooks, and imacs from any IT infrastructure change I will suggest...whats next?


LK


Originally posted by ribitch
a celeron isnt comparable to the G4. Therefore the P4 systems were used. Win XP Home cannot be added to a domain, therefore a centralized authenication server cannot be used, therefore isnt a business class operating system. OS X and XP pro can all have a centrallized authenication server, therefore are an option in a business.

You talk about mac prices being to high for you to purchase a system. So I give you facts to prove they are cheaper than a PC in the long run. So you say well a business never looks into thinsg like that. Your business may not, but when a company has 50 systems, they see the long term benefit more than a short term. Thtas why some companies are ditching CRT's in favor of LCD's. The energy used is much less, offestting the extra expense.

Your "insider" knowledge of a working business is pretty weak. Apparently your companies IT staff isnt really looking at overall costs of a system. Many large companies require this sort of data, apparently yours doesnt.

As far as the benches go, those benches are a few months old. I have seen them before. The mac was on top of the photoshop tests. The 3D tests (Bryce & quake) proved the non mac systems were best for 3D tests. Nothing new. A mac isnt a high end 3D system.



Prices of the PC's were comparable to the prices of the mac 7 months ago. Powermacs are expected to be refreshed on the 17th. DDR or a G5 is expected if any changes are made, and the playing field will be leveled once again.

PC's are updated constantly. Macs on the other hand are updated an almost fixed length interval. (macworlds)

Also note that photoshop 6 was used on the photoshop tests. They say OS X was used, photoshop 6 is not an OS X app. It is using classic mode (an emulator of OS 9). There is a large speed gain from using a native OS X app, such as Photoshop 7. I am curious on how the photoshop tests will fare against the fastest PC if photoshop 7 was used.

[ribitch]

LK - photoshop 6 was running in classic mode. The windows equivilant would be like running windows XP, with virtual PC running win 98 on top of it. XP is a multiuser current OS, 98 is a primarly single user end of life OS. Sure the speed is going to be slower than it would if you were running photoshop in XP, but thats the way the tests were ran. The second set of benchmarks provided by Apex proved there was a big speed difference between enviroments.

You also say that you have all this experience about making decisions for companies. I highly doubt you have ever provided TCO's on PC's and macs to you higher ups. Or how a mac is cheaper to support than a PC (provided the tech has a clue on what they are doing) And fewer technicians are needed to support a given amount of macs compared to PC's. That is partially why more companies dont have macs. As far as power consuption goes, if the facts were presented correctly unbiased to a board deciding on the purchase, one would see macs get much more attention. But again, since people like you have the bias, the information will rarely be presented, and if it was presented, more than likely incorrectly presented. Once again putting the mac at a disadvantage.

Sure there are LCD's being bought by companies for other reasons than TCO, but a large amount of companies do base decisions on this. Compare a 17 inch CRT with a 15 inch LCD. They have about the same size viewable area, but the TCO is by far cheaper on the LCD. Now multiply that by the number installed at one location. Again, huge price savings, which means more money kept in the companies bank account. Thats how business works.

For a home user, power consumption isnt as big of a deal as it is in the corporate world. There are many things that influence a home user. Games are some reasons. Are games the best reason? Not exactly. People say they buy a PC for games, since they have great capabilities and many games are available. Wouldnt it be cheaper to just get a console instead? Well, they do it for online gaming? Isnt that available on consoles? Well they do it because more software is available? Many PC versions of games have expansion packs and updates. Most of these are included with the mac releases of teh games.

Now as far as pricing goes on that dell, what did it include? Did it match the requirements for the comparison that I stated. The included softtware on the PC should be an equivilent type as on the iMac. Was a DVD authoring app included? Or video editting? Or photo? What about optical mouse? 56K? 10/100? firewire(may omit if not available)? Office type suite?

Also, were the rebates taken into account for the final price? No mail in rebates can be used as stated. Its the original price paid that is being determined. When you can configure the system to match the specs, post the specs and a price.

[LegendKiller]

See, what you dont understand, is those figures for "support" only include a VERY small sample. I would bet that if you asked a larger sample of the tech support population you would get MUCH less in the way of hours spent on fixing computers.

You said something around 11 hours per PC per year. With a company the size of mine, which has about 450 computers in my building, with ONE, I repeat 1, tech support person, she would have to work 13.5 hours per day for 365 days in order to meet the needs of the 450 computers. Now, for some reason, I do not think she works 13.5 hours, weekends, holidays, and even her birthday. She even leaves early 3 days out of the week to pick up her kid...wow, thats pretty amazing for somebody who is supposed to be working on a computer all day.

Now, as far as your TCO figures go, companies are switching to LCD's because they save FOOTPRINT space and are easier on the eyes. You mentioned before that you are still in school, probably undergraduate, you might not know everything that goes on inside of a company, I dont either, but I know quite a bit.

Power savings ARE a concern among financial managers, but if the company were to look at 1. The cost of retraining people on a mac, 2. The loss of productivity for switching to a mac 3. THe compatibility with legacy systems, and the costs of upgrading those to new systems which the mac would work with 4. The initial setup costs (getting IT people who are used to working with macs), I can go on and on and on. It just doesn't make sense. Especially if you are going to save around 10k per year on power consumption, thats NOTHING. Show me 1m per year in savings and I might start thinking about it, but even then, thats less than .01% of a large companies annual revenues.

You have proven before with your figures about "salvage" costs that you have no clue about the real financial world and its consideration on IT spending. Hell, I have only scratched the surface in the past 2 years. I DO know how decisions are made and if I thought that the mac was a VIABLE alternative then I would present it to higher ups. However, what you dont understand is that if I DID put it in there, and it is NOT viable but I THOUGHT I should tell them, then I would be out of a job rather quickly because I would be wasting their time.

THis debate will really go nowhere, I mean, seriously, it is obvious that apple has your captivation, totally. Thats all good and fine with me. I do not give allegiance to any one company, I look for the best deal in a rational sense, categorizing priorities. We just have different priorities. Most of mine are usually correct and that is why I have the job I have, but I am not saying yours are incorrect either.

I had a friend who was a rabid mac lover. He was a very good friend in my undergraduate days. I would rip my PC apart and put some newfangled part in, and if it wouldn't work he would have a few hundred laughs on me. I would joke back with him and he would become offended. He was a person who could give a lot on the topic of PC's but couldn't take it on the topic of Macs. We hardly talk to eachother now.

Thus, I will no longer discuss this topic. However, I would expect you to also honor this, and maybe even tone down your level of condescending comments on here, sometimes they are only inviting problems inside of the GotApex community.

LK
Shawn

[sbp]

And to think people thought jase71 and I were bad in going on and on in threads. :heh:

[Apex]

Originally posted by sbp
And to think people thought jase71 and I were bad in going on and on in threads. :heh:

Their badness in no way takes away from your and jase71's badness.

[Ladogaboy]

Originally posted by Apex


Their badness in no way takes away from your and jase71's badness.

:stupid: in a big way...

:D