Replies to post #130391 on Intel Corp (INTC)

[chipguy]

Which is why they put the long disclaimers in the small print below.

The small print typically describes what is being compared and
how. That puts Intel miles ahead of more than a few companies
I could name in terms of supporting performance claims.

The other poster named Geekbench because it is such a amateurish
effort that some of the component programs performed FP operations
on data taken from uninitialized memory which contained a lot of
denorms which unfairly penalized hardware/OS combos (x86) set
up to handle such data rather than flushing to zero (typical ARM).
These clowns only fixed this in rev 3.0. I wonder what other crap
still lives on in their code.

Contrast that to SPEC CPU in which candidate programs are carefully
vetted by experts from many different hardware, software, and compiler
vendors as well as knowledgable computer users from government labs
and universities to ensure consistent and reliable operation over a
range of architectures.

It is like the difference between drunk rednecks playing with fireworks
and a NASA launch.