Author: Uri Blass
Date: 02:14:05 05/27/03
Go up one level in this thread
On May 27, 2003 at 05:00:54, Sune Fischer wrote: >On May 27, 2003 at 03:03:57, Uri Blass wrote: > >>Here is some data about the improvement from 450 mhz to 1200 mhz based on the >>results if I did not do a calculation error >> >>Tiger15 58 >>Deep Fritz7 107 >>GambitTiger2 71 >>Junior7 57 >>Fritz7 110 >>chesstiger14 87 >>shredder5.32 55 >>Hiarcs8 95 >>Rebel Century 106 >>Gandalf5 104 >>Gandalf4.32h 132 >>Crafty18.12 134 >> >> >>You can see that Fritz simply have problems with the 450 mhz relative to most of >>the programs. >> >>The only competition for Shredder in 1200 mhz is Fritz and we can expect >>shredder to lead with big gap in 450 mhz when this program is out of >>competition. > >That is a pessimistic angle you present here I think. >You could say fritz has problems on 450, or you could say fritz is really >optimized on 1200 and all the others have problems on 1200 MHz. > >At some point you probably have to design for the chip. >Do you expect to run on chips with very small cache or very large cache? >Do you want to write optimized assembler for new instruction sets? > >Obviously things like that might influence the way you program. It is possible >Fritz has simply been tuned to take full advantage of modern chips, more so than >the other programs. I do not know. I do not optimize for a special processor and I have no plans to do it in the near future. Uri > >-S. >>Uri
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.