Author: Robert Hyatt
Date: 13:55:04 09/29/99
Go up one level in this thread
On September 29, 1999 at 15:19:34, Dann Corbit wrote: >On September 29, 1999 at 15:05:16, Joshua Lee wrote: > >>I figure Fritz and other fast searchers look 13 or 14ply complete in the >>middlegame and the slow smart programs like Hiarcs look between 8 to 11 ply. >Nope. In the same time frame, they will get approximately the same ply depth. >You must have a whoppersome time control to get to 14 plies, or a really fast >machine. (Or you are considering end-game with a nearly bare board). If you >can get to 14 ply at the opening, I would like to know what kind of a system you >have. > at 3 minutes per move I get to 14 plies all the time on my quad.... but that isn't the same kind of 14 plies that hiarcs would search if it could get to 14 plies, of course. His point was that 14 plies for a null-mover is like 12 plies for a non-null- mover program... >>I have seen some messages that said Hitech looked about 8ply in the middlegame >>it running at 150Knps. If this is so wouldn't most PC software be stronger. This >>doesn't make sense. Hitech was around 2400elo , most programs say they are that >>strong as well but if you look at Rebel 10 it achieved this on super cooled >>cpu's. In reality most computers are under 2400 on regular computers K6-3 >>400's,Pentium III 600's (at rebel's site it estimates the Kyrotech's K6-3 600 to >>be as fast as a Pentium III 900 if it existed), and maybe K7 500. Another comp >>Deep thought i believe looked 10 ply complete, well i think i figured it would >>take over a 4Ghz machine to do this for Nolot pos1 for Hiarcs , anyways what >>middlegame position were people comparing the Ply depth to one with 35 possible >>moves or one with 99(i think that was the highest number seen)????? >Ply depth is not a function of NPS. Fast searchers don't expend a lot of energy >per ply and look at more 'absurd' position. Slow searchers spend a great deal >of energy looking at a node and ignore nodes that look bad. But they will get >about the same ply depth in the same time frame. At least that is my experience >with Crafty/Rebel/Hiarcs.
This page took 0.01 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.