Got|Apex? Review: VIA KT333 Mainboard Comparison: ABIT AT7 MAX vs. ASUS A7V333 6/6/02

[Leon]

Thanks for reading the review. Please post any questions or comments in this thread.


http://www.gotapex.com/reviews.php?rev=mainboards/via333-comp/index.html

[Tcheshire]

one thing i found rather out of the norm with this review.

1.) the RAID scores are nowhere near what they should be with tweaking. i must confes i don't own the at7 but i do own the AX7-333R, and my RAID-0 Sandra score is ~45000, to get this type of score you HAVE to change the computer type in device manager to Standard PC instead of any type of ACPI Computer (HighPoint must have an ACPI irq sharing problem, even if there's no card in the shared PCI slot).

Would have been nice to see the memory timings you guys used as well as the Make of DDR (Crucial etc.)

Just my $0.02

Tcheshire

[sbradfor]

Thanks for the review. I've been trying to decide between the ASUS A7V333 and the AT7. For the last two weeks, I've been heavily researching these two boards, both in the newsgroups and on the web. Your review went into more depth than most of the other 30 or so reviews I've read on each board. I really appreciated this.

I was surpised by your RAID performance findings. Both in the newsgroups and on the web, people are complaining about the RAID performance of the A7V333 and likewise how much faster the AT7's RAID is. Did you get the scores reversed perhaps? You're the first person that I've talked to (or review I've read) that has found the Promise controller to outperform the High-Point. Would you please post exactly how you conducted the benchmarks, what the scores were, and what drivers were installed (and in what order -this can make a difference)? I would like to see why you're getting results that are different than the vast majority of other people I've talked to that have compared these two boards head to head.

You also didn't mention whether or not you'd turned the FSB speed down to spec on the A7V333. ASUS runs it slightly higher than the spec (spec is 133 and ASUS runs it at 135 to give it a jump in the benchmarks). If you didn't turn it down, this could be why the ASUS was faster in some of the benchmarks. This is worth noting. You can get more info at: http://www.tomshardware.com/mainboard/02q2/020509/kt333-11.html

Many people have commented on the A7V333 having problems with fluctuating voltage. Did you test for this or notice it?

With the AT7, many people have had problems with the NIC. It seems to cut out after about 30 minutes of video streaming. Did you test this?

Thanks for the review. I'd love to hear more, especially about the RAID questions I asked above.

Shannon

[DARTH]

For the RAID performance I too was shocked. In fact when I first received the ASUS board I commented to ASUS that they probably should have gone with the HPT374. For testing of the RAID performance:

The A7V33 had a 20GB Maxtor 7200 RPM drive on each of its channels.
The AT7 had the same drives, one on the "3rd IDE Port" and the other on the "4th IDE" port.

Drivers used: Drivers that came with each mainboard

Driver installation: Both drivers were installed at the time of the WinXP install as that is necessary to be able to install windows.

The memory used was Mushkin PC3200, the memory bus was 166MHz so the additional 2MHz that ASUS adds at default should not be a factor. And if it were run at default the 2MHz should not yield that kind of a margin of difference between the two under Sandra.

I do not have the proper tools to take exact reading of voltage, so I would not considered any findings I could report publishable (sp?) .

Once I get my AT7 replaced, the test unit kicked the bucket, I will try the "change the computer type in device manager to Standard PC" suggestion and see if it turns in better marks.

Thx for reading guys,
Stephen

[JYE]

I am trying to decide between the Asus and the AT7. Currently, I am running an Epox 8K7A. The Epox board has a keyboard power on feature. When the power is completely turned off, I can boot my system by pressing any key on the keyboard. I want this feature in my next board. According to the ASUS manual on their website...the A7V333 supports the keyboard power on feature. Can you confirm whether or not the Asus board really does this? Also, given that the AT7 is legacy free, does it support the keyboard power on feature?

Thanks

[Tcheshire]

That's another funny thing about winxp.. last install i did with xp it replaced the drivers i used when installing with it's own AFTER i got XP installed, so you may want to use the ones off the disk and install them again after you get xp up and runnin', i always use new drivers but for reveiwage i can see how you'd like to use the ones that came with the board :)

Was the RAM set for CL2 or 2.5, 4 way interleave etc.? i wish i could get my hands on some crucial from one of my wholesalers here in Canada :(
The stuff i'm using has Samsung chips but it's made to run 166 and that's about it, i can't even run 145 FSB (145+33 RAM) stable at lowest mem timings :/

Tcheshire

[shadco]

The results presented in this review are entirely inconsistent with mine.

I can't put it all in one post so I will cut it up into 3 sections

I own both boards and will post the results I have measured.

All results are at 185 fsb

Mem timings on the Asus are 2-2-2-5 1T

Mem timings on the Abit are 2-2-2-5 2T

Sandra Membandwidth

A7V333

http://mywebpages.comcast.net/ddr333/A7V16/1006185.jpg

AT7
http://mywebpages.comcast.net/ddr333/AT7/at7sanmem185.jpg


Comments?

regards

Shad

[shadco]

Mem timings on the Asus are 2-2-2-5 1T

Mem timings on the Abit are 2-2-2-5 2T

Sandra Membandwidth


A7V333 3DMark2001 SE Abit gf4 ti4400 300/645
http://mywebpages.comcast.net/ddr333/A7V16/3d11451.jpg

AT7 3DMark2001 SE Abit gf4 ti4400 300/645
http://mywebpages.comcast.net/ddr333/AT7/3D11819.JPG



Comments?

regards

Shad

[shadco]

All results are at 185 fsb

Mem timings on the Asus are 2-2-2-5 1T

Mem timings on the Abit are 2-2-2-5 2T



A7V333 ATTO Diskbench 32/64 Fat32
http://mywebpages.comcast.net/ddr333/A7V16/attoprm1.JPG

AT7 ATTO Diskbench 16/16 Fat32
http://mywebpages.comcast.net/ddr333/AT7/attoat7.JPG

As a final I have an HPT 374 Rocket Raid controller installed in the A7V333 now and here is the ATTO 16/16 Fat32 with the Highpoint Controller
http://mywebpages.comcast.net/ddr333/A7V16/attoavr404.jpg

I don't know how you got your results but the AT7 has been consistently faster at all settings.

Comments?

regards

Shad

[Tcheshire]

Holy !@#$ i wish i could run 185 FSB heh.. stupid ram...

[DARTH]

WOW!!! Smoking rig bro... Well for one I could not run either board on a 185MHz FSB. Another thing is that my Mushkin PC3200 could not run at those aggressive memory timings...In terms of RAID I have already listed my steps for installation and configuration etc...

I do not see how your results even compare to mine...Different system, your system is completely overclocked...even the graphics card...Which drives did you use for testing etc...So you can not compare your systems test with mine...But if you found the AT7 to be a better board then I am stoked for you...As I stated in my review, I had to go through 2 boards cause they both were quite flakey...I am awaiting another AT7 to see if that fixes things...Bottom line is the everyone is entitled to their opinion and I liked both boards..The AT7 got a 9 afterall...the one real seperation factor was the stability of the ASUS versus the issues we had with getting a working AT7.
Thx for reading,
Stephen

[shadco]

Well for one I could not run either board on a 185MHz FSB. Another thing is that my Mushkin PC3200 could not run at those aggressive memory timings

That surprises me since I'm just using plain old Samsung PC2700 modules and the facilities built into the mobos. These are the settings I use for 7/24 ops and both of these boards are rock solid.


I do not see how your results even compare to mine

185 was the only setting where I had a complete set of comparo's available for both boards at the same settings. At any setting the AT7 held a similar edge over the Asus whether it was 133/166 default or maxed out.


Which drives did you use for testing etc...

I used Maxtor DX740's all the way around. The Atto results for the AT7 were done the first run using 16k stripe/16k cluster and FAT32. The Asus was tuned to death to get those results. Part of the problem is the crippled functionality of the built in Promise lite controler which only supports 64k stripes and I had to use 32k clusters to even get close to the AT7. Both systems had the Via raid patch 1.05 installed and the Asus had PCI 2.1 Latency enabled in the bios. With out that the writes on the Asus were down to 30Mbs. The Raid controller is the only area where I have a major problem with the Asus board it is so bad I installed a Highpoint Rocket raid card which is the functional equivelant of what's built into the AT7. It's really kind cool because it allows you to install a boot drive in a separate IDE channel thereby allowing independant IDE channels for DVD rom, CDRW, Boot IDE drive and 2 drives in a Raid 0 array.

I had no problems at all with either board except for being sure that I had the latest drivers for everything.

The Asus does have a nice solid aura and allow's for higher vdimm and vcore voltages without having to do mods. Unfortuneatly it is soft in the performance department when compared to Abit's KT333's. That doesn't make it a bad board at all and it's the board I use for the family machine. If Asus were to switch to highpoint like Soyo has They would definately be considered for my next upgrade of the families machine.

The AT7 absolutely blew me away with it's performance. I didn't expect it to rock the way it does. It is actually more stable at the settings I posted benchies at than the KX7 333R I'm testing out now and it is just as fast.

Later

Shad

[sbradfor]

Darth, when you do get another AT7 could you try loading the latest drivers from each manufacturer and test RAID again?

Your results are just the opposite of what everyone else seems to be getting. Granded, you're running different memory than Shadco and he is overclocking, but there are many people that have remarked about how much faster RAID is on the AT7 when compared to the A7V333.

Also, you said that the Mushkin PC3200 memory you used wouldn't overclock to 185 FSB. Is the Mushkin memory really that poor? I've read several reviews of Samsung PC2700 being overclocked to 200 FSB or very close. I'm just learning about this new memory, but wouldn't PC3200 be at least as good as PC2700? I know the Samsung is 6ns, so maybe that makes a difference. Like I said, I'm new to this, so I might be missing the obvious. Could this over spec (by over spec I mean over spec for both motherboads since they talk about PC2700 memory for them) Mushkin memory be making the difference maybe?

Shadco, thanks for posting your results and following up to Darth's reply. This discussion is helping educate me. Much appreciated.

I'm not trying to point fingers at anyone. I am trying to understand differences and reasons for these differences. I appreciate your willingness to look at this and discuss it. It's helping me understand things. It's a lot of fun!

Shannon

[DARTH]

I will be contacting ABIT in the morning to get the replacement board and yes I intend to see what is up with the onboard RAID...From what I gather, in general many like the HPT374 better, and as I said in my review, I always have considered High Point controllers better then Promise, thus my suprise at the results.As to the memory I used, the limiting factor was the CPU. The CPU could not operate at that high of a FSB.
Stephen

[sbradfor]

Darth,

Have you had a chance to get the replacement board up and running yet? I ordered my AT7 on Thursday of last week. It should be here in another day or two. If I can be of help (if you'd like me to run some benchmarks or anything, let me know).

Also, have you read the big KT333 motherboard review at AnandTech? It has some very useful information as to why the Highpoint RAID controller on the AT7 is much better (and much faster) than the Promise RAID controller on the ASUS A7V333. In fact, the A7V333 came in last in RAID performance for RAID 0 & the AT7 was second. Overall, the A7V333 never finished a single RAID benchmark in front of the AT7.

Here's a link to the article:

http://www.anandtech.com/mb/showdoc.html?i=1636

I'm not posting this information to nit-pick your review with you. I'm interested in discovering why you got such different results in your RAID testing (I think it's interesting to figure this kind of stuff out). I know you want the most accurate information for your readers too. I appreciate your willingness to look into this. Please let me know if there's anything else I can do to help.

Shannon

[DARTH]

Still waiting on replacement board from ABIT. They are trying to help me sort out the problem as well....
DARTH

[sbradfor]

Thanks for letting me know. Will you post here after you've had a chance to work with it?

Thanks again.

Shannon