11.
great work!
How do the diffent drives stack up in
- performance across the drive (slowest, fastest, weighed average taking into account that 'slowest' is extra punished for each read taking longer too)
- cleanliness in the performance across the drive (spikes, patterns, ups/downs, oscillations)
- true random content with near and far autocorrelation (random blocks, random multiple blocks), to eliminate OS design favouring one or the other type of content and make it truly a shootout between the caching algorithms where they should be handicapped most
- power draw, especially when copying across the same drive
- operational temperature read from the drive, and measured on the outside
Mike
[
Posted by: mike nittmann

|
Date: 11/08/06 04:05:35 PM]