Author: Robert Hyatt
Date: 12:04:08 01/12/98
I have just finished cleaning everything up, relative to the null-move mate extension and hash table refinement discussed here. As a sanity check, I decided to once again run win at chess at one minute per move to see how Crafty would do on my Pentium Pro 200. The best for an "official" version has been 297, with it typically staying around 295 or so. here is the "crosstable" for version 14.4: 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 +------------------------------------------------------------ 1 | 0 2 0 0 0 2 0 44 0 0 0 0 8 0 0 2 | 0 0 0 0 0 0 0 0 0 1 0 9 0 15 0 3 | 0 0 0 0 0 0 0 0 -- 0 0 54 0 0 10 4 | 0 0 0 0 0 0 0 0 0 0 0 0 0 4 0 5 | 0 0 0 0 0 0 0 2 0 0 0 0 0 49 0 6 | 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 7 | 0 0 0 0 0 0 0 0 0 0 0 0 3 0 0 8 | 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 9 | 0 0 2 0 0 0 0 0 0 0 0 24 0 57 0 10 | 0 3 0 0 0 0 1 0 0 1 0 -- 0 0 0 11 | 0 0 0 14 8 0 1 0 0 0 0 0 0 0 8 12 | 0 0 0 0 51 0 0 0 0 0 0 0 5 0 0 13 | 0 0 0 0 0 0 0 0 0 0 1 0 0 0 33 14 | 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 15 | 0 0 14 0 0 0 0 9 0 0 0 19 0 0 0 16 | 0 0 0 0 0 2 0 0 0 12 0 0 12 0 0 17 | 0 0 0 0 0 0 0 0 0 0 0 8 0 1 16 18 | 0 0 0 0 0 0 0 0 3 0 0 1 0 0 0 19 | 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 20 | 0 0 0 0 0 4 0 0 22 22 0 0 0 0 0 incorrect: 163 230 sum of times(squared)=46689 The bottom line is Crafty v14.4 misses two (163 and 230) in one minute per move. 163 is solved in 1 minute on an alpha/500, but takes just over a couple of minutes on the P6/200. 230 I don't know about, but I think it will take longer than that, as on the alpha Crafty still did not get it last time I tried. Other interesting trivia. With a limit of 30 seconds, it misses an additional 6 on the P6/200, although on the alpha/500 it still gets these right but does then mist 163. So the simple results are: time/move 15 30 60 correct (P6) 287 292 298 correct (alpha) 294 298 299 Which represents the best results I have gotten so far, discounting Cray Blitz of course. Cray blitz solves 297 of 300 in < 1 second (as I have mentioned, that program times to the nearest second and reports times of 0 for 297 of the positions). It takes 1 second (or a little more, but not 2 seconds) to solve 230, 92 and 222. But this test was run on a *big* Cray and averaged about 5M nodes per second for the entire test suite, which burned 32 processors pretty well. :) For the two I miss now, 163 is the easier to solve, and simply needs a bit more speed. Null-move and razoring adversely affect some of these positions, as I have solved some of them quicker in the past. Sorry, but I don't have any times for a PII/300 yet, although we do have one. My P5/233mmx notebook misses two more, as it is some 15% slower than my P6/200. Any other results to compare with? Be interesting to see who solves what quickly and who has trouble with what. I'm going to experiment with the chess middlegame suite Bruce mentioned. The only thing I don't like is the length of time it takes to run a suite that large...
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.