Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: How is hardware "programmed"? (Technical Article by Hsu)

Author: Georg v. Zimmermann

Date: 12:03:52 02/19/02

Go up one level in this thread


On February 19, 2002 at 11:07:23, Slater Wold wrote:

>On February 19, 2002 at 09:42:57, Gian-Carlo Pascutto wrote:
>
>>Something interesting in that paper is their assumption
>>that each chess position takes about 40 000 cycles to
>>process.
>>
>>Assuming an ILP of about 1.5, and Athlon 1000 would run
>>at only 37 500 nodes per second!
>>
>>I think this is indeed indicative that DB had a huge evaluation.
>>
>>--
>>GCP
>
>Anyone who can read that article, and tell me it's only 50 ELO points stronger
>than a PC with todays top software, is outta their freaking mind.
>
>480 chess processors, backed by 16 RISC workstations, and software that is
>*still* revolutionary.

Hmm then I am out of my freaking mind :)

But of course I know far too little about DB, so the best I have is my "feeling"
aboutit.

Take it easy.

Georg



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.