Author: Bruce Moreland
Date: 10:17:25 06/29/99
Go up one level in this thread
On June 29, 1999 at 07:19:50, Ernst A. Heinz wrote: >On June 29, 1999 at 07:04:13, Vincent Diepeveen wrote: > >>On June 29, 1999 at 06:45:35, Ernst A. Heinz wrote: >> >>>Talking about preparation for the WCCC, I like to tell you a little >>>secret. We were not able to do *any* special preparation because Markus >>>finishes his M.Sc. at the end of July and I focused on writing my Ph.D. >>>during the last months. The version of "DarkThought" that participated >>>in Paderborn was at least 9 month old and is essentially equivalent to >>>the one which played the test games listed on our WWW pages (see URL >>>http://wwwipd.ira.uka.de/Tichy/DarkThought/). The program was mostly >>>out of book after 4 moves or so which is actually not so bad as it may >>>sound at first glance. >>> >>>=Ernst= >> >>Hello Ernst, >> >>I heart you ran on a 21264 chip. How fast is this chip and what >>do you think of it? > >We ran on a 500MHz Alpha-21264 which is about 10% faster for "DarkThought" >than the 767MHz Alpha-21164a we had in Paris. Hence, our hardware speed >hardly increased at all since the Paris WMCC while that of all PC programs >at least doubled since then. > >The Alpha-21264 is a nice CPU -- and it will be even nicer as soon as >the bit-find and population-count instructions are implemented on-chip >in the next stepping of the CPU (EV67/EV68). Performance-wise, the 500MHz >Alpha-21264 is not much faster than 500MHz/600MHz P-IIIs, K6-3s, or K7s >on SPECint95. > >=Ernst= The supplied machines in Paris were 200's, some had 233's or 300's. My program goes as fast on a 533 mhz 21164 as it would on a PII/400. bruce
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.