Author: Robert Hyatt
Date: 07:12:59 09/19/99
Go up one level in this thread
On September 18, 1999 at 22:49:20, Christophe Theron wrote: >On September 18, 1999 at 20:54:46, Robert Hyatt wrote: > >>On September 18, 1999 at 17:01:28, blass uri wrote: >> >>>On September 18, 1999 at 15:40:08, Robert Hyatt wrote: >>> >>>>On September 18, 1999 at 11:42:38, blass uri wrote: >>>> >>>>>On September 18, 1999 at 10:01:04, Masciulli Gianluigi wrote: >>>>> >>>>>>I was reading an article by prof. Giovanni Righini (university of Milan) titled >>>>>>"L'eredita' di Deep blue" on the italian chess monthly "Scacco!" >>>>>> >>>>>>I almost falled from my chair when i red: >>>>>>".... >>>>>>E' da anni che il software scacchistico non migliora piu' >>>>>>..." >>>>>> >>>>>>(trad. No step forward in chess software in the last years) >>>>>> >>>>>>I want say that Righini is one of best italian expert in CC, and he in the >>>>>>article esplain that the progress in the last years is mainly by hardware. >>>>>>So I don't want to contest anything to prof. Righini but im just puzled in >>>>>>this may be true or not. >>>>> >>>>>It is not truth that there is no step forward in chess software in the last >>>>>years. >>>>> >>>>>There are always new and better programs. >>>>> >>>>>Genius3 was the best program some years ago and Genius did no step or almost no >>>>>step forward but other programmers did steps forwards and now their programs are >>>>>clearly better relative to the past. >>>>> >>>>>Junior5 is clearly better than Junior4.6 and the next version of Junior is going >>>>>to be clearly better than Junior5 >>>>> >>>>>The same is for Hiarcs7.32 relative to Hiarcs6. >>>>> >>>>>Uri >>>> >>>> >>>>but the improvements are definitely _driven_ by hardware advances. Go run >>>>the latest hiarcs on your old 386 against Genius 2 and see what happens... >>> >>>I think that before 1997 the main improvements were hardware improvement >>>because the best commercial program in 1994-1997 was genius and there was no >>>significant improvement in genius. >>> >>>I think that in the last 2 years half of the improvement is hardware and half of >>>the improvement is software. >>> >>>I guess that if you give the best commercial program of today(probably >>>Hiarcs7.32) the hardware of 2 years ago(PII300 the hardware of many participants >>>in WMCCC) and play it in 40 moves/2 hours against the best program of 2 years >>>ago on pIII550 the result will be close to equal. >>> >>>Uri >> >> >>I wouldn't disagree... but if you back the hardware up a couple more years, >>Hiarcs will get into trouble.. because it will be searching so shallowly, its >>positional skills won't be enough to fend off programs like genius... >> >>Faster hardware has let us add things to the programs and maintain some minimum >>search depth, things that we couldn't do on slower hardware because the depth >>would drop enough that we would get killed tactically... > > >I think a tournament on slow hardware (386-16 to 386-40, or 486 <=33MHz) >including top programs of the early 90s and programs of today would be very >interesting. We could amongst other things see if there have been really no >advances in software in the last 10 years. I don't think the test would be 'fair' if that is what you want to examine. IE many of the software improvements come as a result of hardware changes that make the programs fast enough that they can do things they couldn't on slower hardware. IE would you play _anyone_ if you could only do a 5 ply search? Would you even think about null-move R=2? Would you have your program spend 50% of its time in the evaluation? THose are all decisions I had to address and the answer would be different if I was doing 1K nodes per second on a 386/16.. > >It looks like we have a lot of people out there ready to organize "at home" >computer tournaments. It shouldn't be very hard to find 2 (or even one) 386, or >even one slow 486, and let programs play on this hardware. > >My bet is that the results would be much more interesting -and maybe more >surprising- that the Xth tournament between Fritz5.32, Hiarcs7.32, whatever7.32, >and so on. > >Volunteers are welcome! I'll be glad to provide you with the latest Tiger for >this event! > > > > Christophe > > > > >>IE crafty plays _badly_ on very slow hardware when compared to programs like >>genius on the same hardware. On fast hardware it does just fine... All >>because of the basic assumptions I make in my search, and the things I do in >>the eval...
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.