opteron VS Itanium 2 (Benchmark cheating)

Trent Piepho xyzzy at speakeasy.org
Fri Oct 31 14:36:09 EST 2003


On Fri, 31 Oct 2003, James Cownie wrote:
> > From way back with early linpack, this has left many benchmarks
> > susceptible to vendor manipulation -- there are cases on record of
> > vendors (DEC, IIRC, but likely others) actually altering CPU/memory
> > architecture to optimize linpack performance because linpack was
> > what sold their systems.
> 
> This certainly applied to some compilers which "optimized" sdot and
> ddot by recognizing the source (down to the precise comments) and
> plugged in a hand coded assembler routine.

Nvidia and ATI have recently done similar things, where their drivers would
attempt to detect benchmarks being run and then use optimized routines or
cheat on following specifications.  Renaming quake2.exe to something else
would cause a large decrease in framerate for example.


_______________________________________________
Beowulf mailing list, Beowulf at beowulf.org
To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf



More information about the Beowulf mailing list