Author: Tord Romstad
Date: 03:23:56 11/23/03
Go up one level in this thread
On November 23, 2003 at 03:38:57, Alastair Scott wrote: >On November 22, 2003 at 16:03:19, ERIQ wrote: > >>On November 22, 2003 at 15:47:54, Jason Waugh wrote: >> >> >>>Up to 20 times faster! 400 ELO stronger! >>we will see ? > >Well, on my Zire 71 (which is almost the slowest ARM PalmOS machine) it is a >factor of 8-10 faster - usually 30,000-40,000 nodes per second as opposed to >3,000-4,000. It's still quite a bit faster than my Tungsten T (original version). The ARM engine does around 25 kN/s in the middle game, while the 68k engine does around 1500 N/s. These numbers are for beta 2 (I haven't downloaded the final version yet), but I guess the final version is no faster. I'm extremely impressed by the new Genius so far. In particular, it is tactically very fast, considering the slow hardware. >The strength is, as usual, hard to quantify but I'd say that games between me >and it are, usually, 1/2 or 2/3 as long as with the old engine ;) I played a 6-game time handicap match between Chess Genius on my Tungsten T and Gothmog 0.4.2 running on a Pentium IV 2.4 GHz. I gave Genius 15 minutes and Gothmog 1 minute per game. This time handicap is not nearly enough to compensate for Genius' hardware disadvantage; I think the PIV should be at least 100 times faster. Genius managed to score 1.5 points (3 draws) in the 6 games, which must be considered a remarkably good score. Of course more games are needed to make any accurate conclusions, but Genius is clearly very strong. Tord
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.