I ran across a fascinating analysis of current-generation SSDs from the point of view of reliability.

http://lkcl.net/reports/ssd_analysis.html

The author was involved in some large-scale deployments of SSD drives. Over 50% of the OCZ drives suffered data corruption, and virtually none of the Intel SSDs did. The author set out to create some tests to evaluate drives with, and found out that OCZ went for performance over reliability, in an insane way.

Further investigation then dug up an interesting nugget: it turns out that OCZ apparently had been warned by Sandforce not to enable a switch in the firmware which would result in “increased speed”. OCZ, in their desperate attempt to remain “king of the speed wars” ignored the advice that doing so would result in data corruption. The results correlate with this advice: at higher speeds, data corruption occurs.

OCZ drives actually worked at some point, but OCZ started shipping drives that were somewhat faster in benchmarks, but failed quite quickly. I had two of these drives, one failed within days, so I returned both of them. I would agree with the author’s conclusion:

The OCZ Management deserve everything that’s happened to OCZ. They should have listened to Sandforce: the history of SSDs would have been a radically different story.

It might be interesting to recreate his torture tests and try this on some of the newer drives like the Samsung 840 EVO.