Re: PPro200 vs PPC604 HANDS ON C bench results

From: Bill Broadley (bill@proto.math.ucdavis.edu)
Date: Mon Dec 16 1996 - 04:58:49 CST


In comp.arch John S. Dyson <dyson@freebsd.org> wrote:
> Imagine a P6-XXX running with a chipset that doesn't stunt it's
> performance. The darned Natoma has only about 70MBytes/sec write
> perf (but pretty good read perf.) The Orion is faster than the
> Natoma only with 4-way interleave. Sure wish that Intel would
> produce chipsets that were good enough for the processors.

On my dell p6-200 with natoma chipset:
Function Rate (MB/s) RMS time Min time Max time
Copy: 102.4000 0.2560 0.2500 0.2600
Scale: 102.4000 0.2550 0.2500 0.2600
Add: 112.9412 0.3440 0.3400 0.3500
Triad: 112.9412 0.3490 0.3400 0.3500

Memory bandwidth seems reasonable.

This was on a 36.6 Mb array btw (definitely not in cache).

The box was running gcc-2.7.2 and linux

-- 
Bill Broadley           Bill@math.ucdavis.edu               UCD Math Sys-Admin
Linux is great.         http://math.ucdavis.edu/~bill			PGP-ok



This archive was generated by hypermail 2b29 : Tue Apr 18 2000 - 05:23:06 CDT