🚀 go-pugleaf

Robert Salasidis wrote:
> 
> karl unterkofler wrote:
> >
> > In article <4bvd1r$qn3@sloth.swcp.com>, "Harlan W. Stockman"
> > <hwstock@swcp.com> wrote:
> >
> > > sidea guys <sidea8120@sidea.com> wrote:
> > > >i personally noticed a big difference in speed going from a win 95 P120
> > > >16-4x to a PPC 100 (pmac7500) 16-4x, like maybe 15 percent or more. I was
> > >
> > > Could you be more specific?
> > > In what applications, or with what
> > > benchmarks, did you personally see
> > > the difference?  "15 %" makes it
> > > sound as if you have some exact timings.
> >
> > A simple test with Mathematica:
> >
> > Timing[N[Sin[1/2],2500]][[1]]
> >
> >      PowerMac 8500/120, 16M RAM:             1.43 Second
> >      Dell Dimension XPS P120c, 16Mb, 512K:   4.27 Second
> >
> > Timing[10001!][[1]]
> >
> >      PowerMac 8500/120, 16M RAM:             2.32 Second
> >      Dell Dimension XPS P120c, 16Mb, 512K:   4.31 Second
> >
> > Regards, Karl
> 
> Nobody argues that the PPC is faster in floating point. However for most
> users, 99% of the time is spent in integer functions. Integer benchmarks
> show that there is very little diffeence between the 2 processors (and
> at least the x86 camp has a Pentium Pro to upgrade to today).

99%??? gee, lets not exaggerate things.  glad that you have given up the specfp war, 
but give em a little more credit than 1% please. integer benchmarks show little 
difference? last time i looked it was:

Pentium 120 MHz  spec92int=> 140.0  spec92fp=> 103.9
PPC-604 120 MHz  spec92int=> 157.9  spec92fp=> 139.2

and today the news tells us that intel just happened to find a compiler math bug that 
caused overestimation by 18% in spec92int figures. (no coincidence that the 166 just 
came out to raise the bar - next thing you know they'll tell us they found the report 
in a white house next to a stack of time sheets.) 

so instead of an 89% difference the pentium would be 73% of a PPC. but you are right, 
you can now get a pentium pro (making the earlier models amateurs?) for more $$$, and 
requiring a fan big enough to make a thinkpad rise off the table. ah, but not to worry, 
they will remove the cache to knock off $$$ and heat...coarse that will slow er down 
some. the real question is, do you believe intel can pass the next math test for gosh 
sakes? they are zero for two lately.

by the way... the 604e manages around 225 specint92 at 166mhz, at the same time its 
smaller and cooler than the 604.