Bires

WinXP: I benched my drives in FAT32, then converted them to NTFS and benched them again. Then I colsolidated my drives to one 40G and benched again (still NTFS). The following is the data I aquired.

I used Sandra 2002 Pro and Nero's cache test.

FAT 32:
Sandra: Nero:
8G: 18,409 27,700
16G: 19,005 20,800
16G: 13,660 12,500

NTFS:
Sandra: Nero:
8G: 14,873 27,800
16G: 17,052 20,200
16G: 14,791 16,000

One 40G-NTFS:
Sandra: Nero:
40G: 12,320 23,800


So...I'm off to repartition my drive and most likely convert back to FAT32. BTW: Boot time droped by about 8 seconds when I converted to NTFS on the three partitions, but went up about 10 seconds when I consolidated those drives to one 40G drive.

__________________
Five years...

Bires

...stabil..ila...wha? What is that "stabillitee" you speak off?

Seriously, my system almost never crashes, and I haven't , in nine years of PC's, lost something important from a power loss or other HD error.

qmanchu

Mind if I ask what the cluster-size of your NTFS partition is? I've read (but not confirmed) that the optimum cluster size for NTFS is 4K. This might explain why you notice the difference in performance.

Bires

I never checked...I just assumed that it would be 4K, since Partition Magic usually sets what it thinks is the best for my comp.


I'm back using 3 partitions and FAT32 now. If I get some spare time next week, I may convert one of the drives and see if performance changes substantially with different sized clusters.