Re:Things I learnt making a fast website
by computator on Sunday December 25, 2016 @02:55PM
(#13255212)
I agree that the 20-30x greater power consumption is hard to reconcile with the benchmarks I quoted. So, yes, something doesn't make sense; the benchmark is not taking into account something or I don't know what.
> the two numbers you quote are different benchmarks
But I don't agree with explanation. Both benchmarks are "PassMarks" (a proprietary measure devised by that company). I don't see them saying that their PassMarks for desktops are different than their PassMarks for smartphones.
Do you really think we're expending 20-30 times the power for a 12 % compute speed-up?
Just to clarify, the two numbers you quote are different benchmarks. You can't compare them, lest say "oh, they're only different by 10 percent".