Author: Jeremiah Penery
Date: 21:48:11 08/29/01
Go up one level in this thread
On August 29, 2001 at 19:07:33, Derek Mauro wrote: >If DB was "smarter" than today's programs (and I believe you that it was), and >you consider today's programs not to be super-intelligent, why is it that we >haven't been able to make smarter programs? It makes perfect sense that in 4 >years we should have made more progress. Did the DB guys just know a hell of a >lot more than we have figured out, or is it that because of some hardware issue >we just can't implement everything, or something else? I remember seeing in 1997 some analysis from CS-TAL of the DB2-Kasparov games. In most cases, after a long search, CS-TAL had a much closer evaluation/PV to Deep Blue than all the other programs. My guess is that CS-TAL is doing a lot of these expensive types of evaluation things that DB was doing - indeed, CS-TAL knows a lot more about king-attacks than any other program I've seen. But the problem is that its evaluation is so expensive that it is simply searching too slowly (something like 20000 NPS on a fast computer?) and gets beaten by much deeper searching programs (like Fritz, etc.). If CS-TAL was getting as many NPS as Fritz (1M NPS?), I bet it would completely destroy Fritz, because it has a lot more knowledge than Fritz. Before the DB2 match, there was a discussion with Chris Whittington (programmer of CS-TAL) about DB2's evaluation, and how it compared with evaluation of micro programs. I believe Bob asked, "How do you think CS-TAL would play at 200M NPS?", to which Chris replied something like, "It would be impossible for humans to handle." DB was probably tuned much more conservatively, and Kasparov was very careful about tactics, or I believe DB may have simply blown Kasparov off the board tactically. But of course this is only my guess. Jeremiah
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.