Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Why is P4 less efficient than Athlon (or P3) for chess programs ?

Author: Anthony Cozzie

Date: 17:11:59 07/03/03

Go up one level in this thread


On July 03, 2003 at 15:02:55, Joachim Rang wrote:

>On July 03, 2003 at 12:27:20, Ernest Bonnem wrote:
>
>>Does anybody here know the (somewhat detailed) technical reason ?
>>Chip architecture ?
>>
>>On the average, for chess programs, a P4 running at frequency F is equivalent to
>>an Athlon at frequency F times 2/3.
>>Actually, it seems that this 2/3 ratio is no longer as bad for hi-end P4 (3.06,
>>3.0 GHz) or the Mobile Pentium : any confirmation and reason ?
>
>The main reason is, that Athlon and P3 have 9 instructions per cycle and P4 has
>only 6.
>
>Hyperthreading and higher FSB may improve the newest P4s.
>
>regards Joachim

Chance of an Athlon ever retiring 9 instructions in one clock cycle:
0.00000000000000000001%.

The main reasons the Athlon gets more instructions per clock (IPC) are its
larger L1 caches and its smaller branch misprediction penalty.  The PIV was
designed to do DSP :)

Anthony



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.