the right way to run this is one cpu, then two, etc. Then you can divide
the times and factor out the optimizing noise. And it _definitely_ needs
to use more than a fraction of a second of cpu time. Bad cache layout
can cause fluctuations more than the total run time for .01 second runs.
Robert Hyatt Computer and Information Sciences
[EMAIL PROTECTED] University of Alabama at Birmingham
(205) 934-2213 115A Campbell Hall, UAB Station
(205) 934-5473 FAX Birmingham, AL 35294-1170
On Fri, 7 May 1999, Andy Poling wrote:
> On Fri, 7 May 1999, Roland Orre wrote:
> > First I have observed that some people have used different optimization
> > parameters for gcc, Terry Alden suggested -O but some have used -O6,
> > which on my machine gives about twice the performance. For this test
> > to give the desired result it is of course essential to have the same
> > optimization setup.
>
> I also noticed that most people are reporting times of small fractions
> of a second.
>
> I'd recommend a longer-running benchmark to increase the signal-to-noise
> ratio...
>
> -Andy
>
> Global Auctions
> http://www.globalauctions.com
>
> -
> Linux SMP list: FIRST see FAQ at http://www.irisa.fr/prive/mentre/smp-faq/
> To Unsubscribe: send "unsubscribe linux-smp" to [EMAIL PROTECTED]
>
-
Linux SMP list: FIRST see FAQ at http://www.irisa.fr/prive/mentre/smp-faq/
To Unsubscribe: send "unsubscribe linux-smp" to [EMAIL PROTECTED]
