One thing I love is people making up random bullshit to discredit Geekbench, like:
"The [Geekbench] benchmark code is more like mobile application code than it is desktop code."
What does that even mean? Code is code is code. Practically all Geekbench does is run integer/floating point calculations and some encryption/compression algorithms. There's hardly any "application code" to speak of at all, it's just making the processor do raw math. The only thing that may throw this off from processor to processor is how it gets compiled.
Of course, their source for this is a tech businessman and analyst with, as far as I can tell, no programming or hardware expertise. He's a marketing guy.
It is fair to say that, for instance, A* processors score disproportionately high on SHA1/SHA2 calculations due to hardware acceleration, and that can skew the overall score. But the same could be said of some Intel chips with AES, and even looking at individual tests, ignoring SHA and AES entirely, A9X is very competitive. And of course synthetic benchmarks are not necessarily indicative of real world performance, but they are still useful overall for gauging processor performance, otherwise no one would use them.
