Based on Ao1's posts about NASPT analyzer in the hiomon thread (which apparently uses data that cannot be easily compressed by Sandforce SSDs), I guess that Anand's and StorageReview's traces use data that is easily compressed by Sandforce drives. I think that is why Anand's and StorageReview's traces seem biased in favor of Sandforce SSDs.
I'm not sure whether Tom's traces are easily compressed by Sandforce SSDs. They seem less biased towards Sandforce drives than Anand's traces.




Reply With Quote
Bookmarks