Benchmark Results: PCMark And Sandra
Although many folks use synthetic benchmarks as a primary metric for performance evaluation, we put our faith in real applications whenever possible. Readers who compare test results between different online sources should find the specific results more useful, though no more revealing.
PCMark shows that the only real difference in performance is small, and primarily limited to drive throughput. Differences of less than 10% are usually due to drive conditioning, adding little to the discussion.
Large difference in Sandra 2011 would have most likely been caused by configuration issues or excessive overhead for a component. Fortunately, that level of deviation does not exist.