Author: Dann Corbit
Date: 13:21:09 04/17/00
Go up one level in this thread
On April 17, 2000 at 15:53:00, Will Singleton wrote: [snip] >Did you do anything to Amy in terms of eval? Or was your contribution to get >smp debugged? I did nothing for eval. I did a port of Pthreads to Win32 I did a port of sleepycat to Win32 I connected the three. Upon profiling, I found that the big bottleneck was in bitcouting and finding the first bit. I did a tweak to the bitcounting which gives a small boost. I made a try at the find-first-bit routine, but my assembly version does not provide the answers that the calling routine wants, so that alteration is on hold now. But I have done nothing with the eval and have no plans to do so. 99.99% of the code is from Thorsten, which is as it should be. Mostly, I just provided interfaces to the missing pieces and a bit of porting work. If a much faster 'first-bit' routine were provided, it will be a giant boost for Amy. Wait until you see another new star. There is a program called Gullydeckel 2 which I predict will be a marvelous blitzer. For some reason, the author is very modest about the program's performance, but my tests indicate that it will be rather strong. Fascinating approach that does not use 64 bit bitboards, and yet hits 100-300K NPS when analyzing on a wimpy PII 300 MHz.
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.