Author: Mike Byrne
Date: 16:56:37 01/20/04
Go up one level in this thread
On January 17, 2004 at 12:53:11, Christophe Theron wrote: >On January 16, 2004 at 22:39:57, Mike Byrne wrote: > >>On January 16, 2004 at 14:08:49, Christophe Theron wrote: >> >>>On January 16, 2004 at 01:14:50, Mike Byrne wrote: >>> >>>>On January 16, 2004 at 00:41:45, ERIQ wrote: >>>> >>>>>in ssdf or fide. >>>> >>>>on a 400 Mhz Palm -- it is my belief that it is near 2450 ( I would say that teh >>>>chances are excellent that it will fall in somewwere from 2375 to about 2500 >>>>with my best guess at 2440) ...the pocket pc version will be about the same >>>>....on both machines it can see over 100K nps in certain positions which is >>>>fantastic for handheld device... >>> >>> >>> >>>You seem to be very impressed by the NPS, but you should know that in computer >>>chess NPS is definitely not a measure of strength. >> >>I agree with you. >> >>> >>>As a matter of fact I could make a few quick changes that would double or triple >>>Chess Tiger's NPS. But the program would be weaker then. >>> >>>Maybe a high NPS was impressive in the '70. Now it's obsolete... >>> >> >>I only mention nps because of Genuis NPS on non arm / non xscale machines >>compared to Genius nps on arm/xscale -- it is quite impressive when comparing >>apple (Genius older processors) to Apples ( Genius code for new proccessors). >>NPS is only woth mentioning when comparing identical program to identical >>program running on perhaps different processors with different optimizations. > > > >You are not comparing apples to apples. > >1) older versions of Chess Genius for Palm did not show the nps. So you cannot >compare to that. Even if you could, you are comparing a 68K assembly of 1987 >with a C code of 2004. The programs are probably very different. I was provided a special copy of the older pocket pc version that displayed nodes. I', sure Richard has the programmer copy that displays nodes for the older 68K assembly. > >2) as far as I know, only the latest version of Chess Genius for PC displays the >NPS. I don't have this one, but I have older versions that did not display it >anyway. The PC version of Genius is, as far as I know, written in x86 assembly. >And it is a totally different program that the one (writen in C said Richard) >that runs natively on Palm. They are probably not searching the same tree at >all. > >So there are things that you cannot compare because of lack of data, and there >are things that you should not compare because they are totally different >engines. I was comparing pre-optimized version of the Chess genius for the Pocket PC to the optimized versions - the programs were exactly the same in all other respects. You don't have to argue with me about this - you can ask Richard directly yourself. > >That makes any consideration of NPS in this case even more useless than it is >usually. > Since you do not have the facts correct, this whole discussion is useless. Michael > > > Christophe > > > > > >>For example on this position (link below) , running a Dell Axim (oc to 600 Mhz) >>it will see exactly the same number of positions in just 31 seconds or 103K nps. >> >>http://www.chessgenius.com/palm/faq.htm >> >>I would to love see CT optimize for the Pocket PC - any chance of that >>happening? >> >> >> >> >>Best, >> >>Michael >> >> >>> >>> Christophe
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.