1518 with two WD400JB 8MB cache, Promise TX2000 controller. Striped with 32k stripes.
Performance depending on file size (faster with large files, slower with small files) depends on the stripe size, which determines what should be considered a "small" file and what is "large".
With my Seagate Barracuda ATA IV on the nforce2 controllers, I was getting about 870 I think.
If your performance is VERY low, check whether the cache is properly enabled for one thing. After setting up my array and installing the Promise management utility, it showed the array as having the drive caches disabled, even though the manual said it should be enabled by default. Also check whether Windows has the cache enabled, which shouldn't be changeable (greyed out). Another thing that seems to sometimes improve performance is making the disks dynamic in Disk Management, but make sure it's something that won't backfire on you before you do it; XP has issues with the way cache works that was resolved for basic disks, but not dynamic disks apparently, but it improves performance at the possible expense of data integrity.
I'd say 1149 with two 5400 RPM drives is probably pretty good.
I'd like to know how some of the scores on the FutureMark results listings were so very very inflated. 13000+? 700MBps read speeds? There's no controller in existance that can read that fast no matter how many drives are simultaneously transferring, and even 2 Ultra320 controllers combined can't get that high in real throughput, let alone whether any system bus could transfer that much data.
Facebook
Twitter