Computer Chess Club Archives

Search

Terms

Messages

Subject: Mach 3.8GHz -- are the SPEC scores believable?

Author: enrico carrisco

Date: 15:21:39 09/28/03

http://www.go-l.com/desktops/machl38/specs/index.htm

A friend sent me the link above for a professionally overclocked P4 3GHz (EE)
overclocked to 3.8GHz.  The SPEC integer and floating point tests are quite
impressive -- a reported +300 over the new Athlon 64 FX @ 2.2GHz.

But using a simple formula to compute what a *perfect* linear speed-up under the
best conditions deems this reported "1749" SPECint score a bit of
sensationalism.

Their own graph shows the P4 3GHz at 1152 on the integer test.

1152 * (3800/3000) =  1459.2

I am curious to know if these scores have been accepted by SPEC.

-elc.

Re: Mach 3.8GHz -- are the SPEC scores believable? Gian-Carlo Pascutto 16:58:49 09/28/03
- Re: Mach 3.8GHz -- are the SPEC scores believable? Jay Urbanski 20:44:07 09/29/03
  - Re: Mach 3.8GHz -- are the SPEC scores believable? Vincent Diepeveen 04:39:43 10/01/03

This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.