vinuneuro wrote:Manufacturer specs are not that accurate then it seems. When I checked them both in this laptop with HD Tune, the Seagate was around 17.3-17.5ms. My Fujitsu is 15.5ms. That's a big difference.
The 7K500 is actually in the 16.8ms range it looks like, just searching google...
Sure, you never know and have no control where the data is placed, or how many fragments. Just stepping one more track is 1.5mSec and maximum latency (sector placed 180 deg from the head) would add 4mSec. I don't usually use HD Tune but I will take a look.
It will be good if you can run Crystal Diskmark then we can see apples to apples comparison.
vinuneuro wrote:...I think there may be a correlation between platter density and performance. The 7200.4 and 7K500 went from 160gb to 250gb per platter compared to the 7200.3 and 7K320; and the read/write speeds went up considerably, but the access times suffered a bit.
Yes, increased areal density is much easier than increasing spindle speed. 5400rpm to 7200rpm is only 33% increase but 160GB/platter to 250GB/platter is a 56% increase. Here is a comparison of high spindle speed/low areal density and low spindle speed/high areal density designs, Hitachi 7K100 vs WD (Apr 15 table):
http://forum.thinkpads.com/viewtopic.ph ... 1&start=60
In the past, data from different heads are interleaved to improve performance. Judging benchmarks from today's single platter and two platter designs, it seems they don't do that anymore.