Username or EmailPassword
"Second, a benchmark cannot be 'wrong' in the sense of numbers itself given numbers are numbers. What one may disagree with is how the benchmark was done (the circumstances). When the very same benchmark were to be done on a similar system, the very same numbers (give it a rounding error, or not) come out of it."
So basically a benchmark is wonderful at telling you how well something rates in the metric you are arbitrarily testing. Where is your scientific data then that proves whatever metric you feel like using actually relates to overall system performance in any meaningful manner?
First rule of benchmarking is that the numbers are only honestly valid for the exact test you are running - an extrapolation is the same kind of hogwash you are condemning here in this review. Why not stop trolling and hiding behind a need for "scientific data" when someone else's opinion isn't egosyntonic to what you believe?