I was just playing with and benchmarking a few Raid arrays last night (all Raid0) and I was wondering if anyone else has had similar findings.
System1: Abit VP6, Dual PIII/1Ghz, Asus V7700 Geforce2 GTS 64M, (2) IBM 30G 75GXP, 1GB PC/133, Win2k Sp2
The harddrives are stripped using the onboard highpoint controller and split to 3 nearly equal partitions. c is fat32, d&e are ntfs OS is resident on E.
system2: Iwill DVD266-R, Dual 1Ghz PIII, Geforce2 Ti 64M, 512M PC/266, (2) IBM 20G 60GXP, Win2k (fresh install no SP)
The harddrives are stripped using the onboard AMi megaraid controller and split to 3 nearly equal partitions. c&d are ntfs, e is fat32. OS is resident on C.
Both systems have VIA 4in1 4.34(final) and the new nvidia Dentonator XP drivers.
-------
results using the drives benchmark on Sisoft Sandra 8.11
System1:
C-- 39,600 (fat32)
D-- 38,400
E-- 38,200
System2:
C-- 24,500
D-- 24,300
E-- 25,400 (fat32)
The numbers varied slightly on multiple benchmarks but thats basically where they were at. I checked and I do have the most recent bios and RAID drivers/bios for both boards.
I originally had dual PIII850/100s in the IWill and the scores were even worse.
The Abit machine (system1) has been running for 8months +/- and has a significant amount of data on it as well as a few programs resident. The Iwill machine (system2) had just been burned in and was fresh.
I guess I expected the 60GXPs to perform alot better. I have (3) 20g 60GXPs on a Win98 machine using a promise Tx2 addin card and they work very well.
Has anyone else seen similiar results (poor performance by the AMI controllers) with the Iwill boards.
Also, why is the Fat32 significantly faster on the access side?