Author: Dan Newman
Date: 13:07:57 07/22/00
Go up one level in this thread
On July 22, 2000 at 10:36:41, Larry Griffiths wrote: >Hi, > >My basic bitboard move generation is in place so I ran a test to get >moves-per-second. It is a brute-force recursive function call that generates >all the moves for a ply; makes a move for that ply; calls brute-force; then >unmoves the ply. I do not have en-passant, castleing, or move validation code >in place. >This ran on a Pentium III 550Mhz Xeon processor... > >Here are my results... > >TCBoard - BruteForce Executed in 1726.16 clock seconds >TCBoard - BruteForce Moves Made/UnMade=3408748153.00 >TCBoard - BruteForce Ply(1.00) Move Count=20.00 >TCBoard - BruteForce Ply(2.00) Move Count=400.00 >TCBoard - BruteForce Ply(3.00) Move Count=8902.00 >TCBoard - BruteForce Ply(4.00) Move Count=197742.00 >TCBoard - BruteForce Ply(5.00) Move Count=4896998.00 >TCBoard - BruteForce Ply(6.00) Move Count=120909581.00 >TCBoard - BruteForce Ply(7.00) Move Count=3282734510.00 > >This divides out to about 1,974,757 moves per second... > >I read a thread in the last couple of months where some folks stated that their >programs were around 5 million moves per second. > >Do I still have a lot of tuning and optimization to do? > >Larry. Mine gets about 1.355 million nodes/s on a 400 MHz Celeron on this test. To compare with a 550 Xeon multiplying by 550/400 might be close enough: 1.3 x 550/400 == 1.86 M/s, which is pretty close to your number. It could be a matter of terminology. Some people were (in some threads) comparing move generation rate in which you call the move generator repetitively without any search. This is generally what I'm talking about if I say "moves per second". Or it could be a matter of faster processors. The above (brute force) test is weighted towards measuring the speed of the make/undo code since you end up calling make/undo about 30x for each move generation--but make/undo speed might be slightly more important than move gen speed anyway... (This test also equally mixes non-capture with capture generation, but the capture generation speed is much more important usually.) Anyway, it looks like you're going fast enough :) -Dan.
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.