Benchmarks: Full Drive Testing
The RevoDrive 350 only suffers a 5% performance loss in our single large file copy test once the drive is full dropping from 629MB/s to 598MB/s.
Oddly, the RevoDrive 350 became 8% faster in the program copy test once full (the 401MB/s result was based on an average of three runs). Truth be told, the results were similar both empty and full, the full test just managed to produce slightly more consistent numbers.
As was the case with the single large file copy test, we find a 5% reduction in performance when using the game data.
The sequential read performance is much the same with the full drive dropping just 4% of its original performance when empty.
What we are looking for is a reduction in write performance caused by the drive having to clear blocks before storing new data. Here we do see a significant drop in performance when measuring sequential write performance as the OCZ RevoDrive 350 drops from 902MB/s to 705MB/s, a 22% reduction.
Although the RevoDrive 350 was much slower full when measuring sequential write performance we find that the random 512K read performance has increased by 8%.
Again, like the sequential write results we find that the RevoDrive 350 is much slower full when measuring random 512K write performance, showing a similar 20% drop from 772MB/s to 615MB/s.
Interestingly when looking at the random 4K-QD32 read results we find that the OCZ RevoDrive 350 is slower, much slower when full, suffering a 25% reduction in performance.
We expected the random 4K-QD32 read performance to remain mostly identical once the RevoDrive 350 was full and that wasn't the case. However, the random 4K-QD32 write performance did remain much the same.