At 01:46 PM 11/23/2001 -0500, brian wheeler wrote:
>On Fri, 2001-11-23 at 13:41, Simon Cozens wrote:
> > On Fri, Nov 23, 2001 at 06:04:29PM +0000, Simon Cozens wrote:
> > > * Rewrite mops.pasm to use integer PMCs, and compare the speeds.
> >
> > I couldn't wait. :)
> >
> > % ../../test_prog mops.pbc
> > Iterations: 100000000
> > Estimated ops: 200000000
> > Elapsed time: 9.948440
> > M op/s: 20.103654
> >
> > % ../../test_prog mops_p.pbc
> > Iterations: 100000000
> > Estimated ops: 200000000
> > done
> > Elapsed time: 20.994231
> > M op/s: 9.526427
> >
>
>I don't get it. What kind of machine was that on? Here are my numbers:
>
>
>[bdwheele@thor parrot]$ ./test_prog mops.pbc
>Iterations: 100000000
>Estimated ops: 200000000
>Elapsed time: 11.016154
>M op/s: 18.155156
>[bdwheele@thor parrot]$ ./test_prog mops_p.pbc
>Iterations: 100000000
>Estimated ops: 200000000
>done
>Elapsed time: 14.288639
>M op/s: 13.997134
>
>My I-regs one is slower than yours, and my P-regs one is faster...this
>is on a P4/1.7G under Linux 2.4.15pre8
Welcome to the joys of architecture specific wackiness. :)
Dan
--------------------------------------"it's like this"-------------------
Dan Sugalski even samurai
[EMAIL PROTECTED] have teddy bears and even
teddy bears get drunk
