Author: Robert Hyatt
Date: 08:22:14 08/19/02
Go up one level in this thread
On August 19, 2002 at 08:28:47, Omid David wrote: >On August 18, 2002 at 23:22:24, Robert Hyatt wrote: > >>On August 18, 2002 at 15:42:34, Omid David wrote: >> >>>On August 18, 2002 at 09:06:02, Jorge Pichard wrote: >>> >>>> Kasparov proved that he can defeat programs at fast time controls when he >>>>defeated Deep Thought in a game/90 two games match in 1989. This program was >>>>weaker than Deep Junior is today, as it searched well over 2,000,000 NPS, but >>>>didn't have as much chess knowledge as Deep Junior. He also defeated Deep Blue >>>>in 1996. This program is obviously much faster than Deep Junior is today, but in >>>>my opinion Deep Junior still has more chess knowledge than Deep Blue had back in >>>>1996. >>>> >>>>PS: It is hard to compare Deep Blue of 1997 vs Deep Junior of today, but in my >>>>opinion Deep Junior Chess Knowledge could make up for the difference of Deep >>>>Blue super calculating power of 1997. >>>> >>>>Pichard. >>> >>>Deep Blue retired at peak, exactly like Fischer. However, saying Deep Blue is >>>the strongest ever, is as ridiculous as saying Fischer is still the strongest >>>player. >> >> >>Maybe or maybe not. If someone says to you "I have program X running on >>secret hardware and it is searching at over 200M nodes per second" what >>would _you_ conclude? Would it particularly matter whether program X was >>gnuchessx or Fritz? IE I saw gnuchess clean the commercial's clocks a few >>years ago on ICC when someone ran it on a _really_ hot box. It rolled over >>everyone, commercial and non-commercial, with relative ease. >> >>We know how fast DB searched. Is there _any_ convincing argument to offset >>that ridiculous speed??? > >Instead of comparing the speed, or algorithmic sophistication seperately, we'd >better compare them altogether, comparing only the *strength*. In 1997 Deep Blue >was most probably the strongest chess playing computer, however its mere >retirement shows that IBM wasn't sure whether it can still hold the lead. > >Regarding your example, if someone tells me his program X running on a secret >hardware is searching more than 200M NPS, I'd suggest him to come to a >tournament and prove its strength (not mere speed!). That is the difference between us. I _know_ what the speed does. Given two equivalent programs, one being 200x faster will be more than simply dominant. DB was not a "bad" program. And with its speed advantage, it was a nearly invincible program, compared to other computers.. > >Deep Blue can very well still be the fastest searcher, however that shouldn't >lead to any conclusion regarding strength. Since 1992 no brute force searcher >has been able to win the WCCC. Take MIT's 256x CilkChess and the 1024x DIEP as >the latest examples. (In fact all the WCCC winners since 92 have been forward >pruners.) So? DB did some forward pruning, as we already know. It had a branching factor that was nearly as good as todays programs, well under 4.0 which is in the right ball-park. I don't see _anything_ that says (a) a forward pruner is better; (b) that DB was a pure brute-force program; (c) that a pure brute-force program, with good search extensions, can't win a WCCC... > >Omid.
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.