Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Prescott/"Pentium 5" (?) FYI

Author: Aaron Gordon

Date: 06:58:22 03/09/03

Go up one level in this thread


On March 08, 2003 at 05:48:23, Tom Kerrigan wrote:

>Some info/speculation on the next Intel core, due out by the end of the year,
>probably at 4GHz:
>
>http://www.chip-architect.com/news/2003_03_06_Looking_at_Intels_Prescott.html
>
>Looks like it will have much higher (I'm guessing ~20%?) per-clock performance
>than current Pentium 4s. That would give it a SPECint of ~1770.
>
>(It will run multithreaded programs much faster, too.)
>
>A 2GHz Opteron gets 1202. If AMD can get their chips up to 2.5GHz at the end of
>the year, that means a SPECint of ~1502.
>
>(I figure 2.5GHz because that's the limit of Athlon overclocking right now, just
>like 4GHz is the limit of P4 overclocking. Opteron might run faster because it
>has more pipeline stages and SOI, or it might run slower because AMD is
>apparently having problems with SOI. So I'd call it a wash and say 2.5GHz right
>now. Notice that these are Opteron scores, not Athlon 64 scores, so the systems
>we'd actually buy will likely be a little slower.)
>
>So my guess is that AMD will be 20% slower than Intel at the end of the year,
>and that's not even counting the (potentially quite large) benefit of
>multithreading. *sigh* Such a shame. If only AMD was executing on Hammer better.
>Might have to sell some shares eventually.
>
>-Tom

Limit of P4 overclocking on air is about 3.5GHz, limit for AthlonXP's on air is
about 2.5GHz. With freon the limit for P4's is ~4GHz and the limit for
AthlonXP's is ~3GHz.



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.