[RC5] Tillamook cpu selection problem

Dave Avery cwax at pacbell.net
Sat Nov 29 14:56:34 EST 1997

Jim McManus wrote:
> I ran the benchmark from the GUI and, no surprise,  the pentium, pentium MMX
> core was the fastest. I was curious how the processor is determined. I noticed
> that when NT starts is somehow knows it's in a "genuineintel" processor. Does
> the code look for something embedded in the processor that is returned when
> some certain instructions are run? Or does it run a few instructions and guess
> based on the results?
> Reid Conti wrote:
> > > I just got a Toshiba 750 which has the new 233MHz mobile pentium, and almost
> > > everytime I start the client it picks a different core. Out of the six  available
> > > cores,
> > > it's picked everyone except the AMD K6. Any ideas on why it keeps cycling the
> > > processor cores?
> >
> > They're probably close.  Try doing -benchmark2 specifying each core.
> > You'll probably find 2 or 3 that have VERY similar speeds, some that are
> > WAY off.  Then -benchmark the last 2 or 3 that are close, and pick the one
> > that looks the fastest :)

it runs a quick benchmark with each of the cores and uses the
fastest tested. if anything else is running it will be mislead

To unsubcribe, send 'unsubscribe rc5' to majordomo at llamas.net
rc5-digest subscribers replace rc5 with rc5-digest

More information about the rc5 mailing list