Unfortunately we need "3rd party" benchmarks to disapprove the Intel haters that it does indeed perform like what Intel says. I know ARM did benchmark their CPUs under the SpecCPU, but the exact numbers were never disclosed. And A9 does have one advantage in paper over Atom, OoOE. So it becomes even harder to believe Intel's claims that Atom has 30-40% advantage over the A9 in a single threaded benchmark.
> To attempt to just dismiss it as a compiler benchmark is a cop-out > of any unpleasant information it is telling you. I would tend to > believe it as a general guide until specific application benchmarks > tell you different.
The positive spin is that ARM has crappy compilers?
Compilers are definitely a selling point for a platform. Intel has spent a ton of money on compiler development and that's part of the platform. Even AMD likes to run some of their benchmarks with Intel's compilers.
Market Data 
Markets 
AI
