Navigate: Advice Beginners BIOS Guide CPUs Links Mainboards Memory Network Storage Video/Sound Cards Contact Forum SiteMap Sponsors Home PC Hardware from Made-in-China.com Search: Enter your search terms   Web www.lostcircuits.com Submit search form	.
Prices:

LOSTCIRCUITS		SHORTCUTS: The Challenge Fujitsus and Promises Assembling the RAID Building the System Test configuration HDTach vs. WB99 ATTO vs. Sandra Back to HDTach, WB99 Network File Copy Conclusions Fujitsu Drives Comments?
Internal 4-Way SATA RAID For The XPC     A Concept Study
(Review by MS, April 20, 2003)

We mentioned earlier that the drives used are first generation SATA adaptations of notebook drives, soon to be followed by the next generation running at 5400 rpm and featuring an incredible 16 MB on disk cache. Keep this in mind when looking at some of the benchmarks here. All benchmarks were run with a stripe size of 32k, which is the smallest size available in a 4 disk RAID 0. With all SATA drives we have tested thus far, small stripe sizes have yielded the best results, regardless of applciation or benchmark used.

This looks not only pretty ugly but also very slow. Each of the drives can sustain sequential reads at the platter OD of approximately 25 MB/sec so why does the RAID setup not show anything higher. Rest assured, this is an artifact of HDTach which is not capable of coping with high sequential transfer rates.

To prove our point, we ran Winbench 99 disk inspection test on a 30 GB partition of the drive and the result is quite different from what HDTach shows.

Up to 95 MB/s sequential reads in WinBench99 are a different story than the one told by HDTach. The jitters at the outermost zone of the drives are caused by the platter transfer exceeding the host transfers causing a cache overflow and momentaneous stalling of transfer.

next page:    => ATTO, SiSoft Sandra =>