There are five big factors in hard drive performance:
1) Spin rate, equal between the drives being discussed.
2) Data density, WD is a bit lower than Maxtor and IBM here, and this translates to higher sustained data transfer rates on some new Maxtors and the IBM GXP180 than the SE WD's produce. Data density also has an effect on seek performance, but is not as directly linked with it as it is with data transfer rates.
3) Seek performance, this is where the moden Maxtors really drop the ball. Seek performance is much worse on the latest Maxtors than even the drives like the preceeding D740X. WD is better here, and IBM better yet.
4) Cache size, WD clearly leads the pack here, and it matters more than anyone gave it credit for before the proof was produced by the SE series dominance over both the 2MB WD siblings and the competion's drives. Once IBM GXP180's become available with 8MB buffers though, their improved data density (which translates into improved data transfer rate) and seek advantage will allow them to clearly surpass the current WD SE.
5) Cache policy and optimization, an area where Maxtor has really dropped the ball recently. The primitive buffer management of the latest Maxtor designs really hampers their performance. Between this and the poor seek performance the newest Maxtors are not the drive to use for speed, even though they have very good sustained transfer rates. This area is a WD strength, with the firmware and buffer on the BB and SE drives outstanding. This factor combined with acceptable (but not leading) seek performance and the 8MB buffer is the winning package. At least until the 8MB GXP180's arrive, although IBM seems inclined to only use the 8MB buffer on the largest size (180GB) and charge a fortune for them.
