Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Floodfilling attacks

Author: Matthias Gemuh

Date: 15:45:45 10/25/02

Go up one level in this thread


On October 25, 2002 at 12:19:15, Gerd Isenberg wrote:

>On October 25, 2002 at 12:05:30, Matthias Gemuh wrote:
>
>>>
>>>How long does your getQueenAttacks take in a loop with 10**9 runs? My
>>>mmx-routine on Athlon XP 2.1+ takes about 51 seconds (no memory access, no
>>>conditional jumps, up to four mmx-instruction in parallel).
>>>
>>
>>
>>
>>"rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"
>>
>>It takes 329 secs on Athlon XP 1.8+ for white Queen(s).
>>
>>Thanks for the link ! I will try the optimizations you propose in other posting.
>>
>>/Matthias.
>



Russel Reagan's routines/macros need 282 secs to compute the Queen attacks.

/Matthias.





>
>not so bad as i thought. But instead of using the rotated lookups this one
>should be runtime independent from the position.
>
>The rotated lookup version (inlined) took about 20 til 55! seconds on my pc,
>with constant occupied bitboard but some cyclic squares (loopCounter&1 was 55!
>seconds).
>
>Gerd



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.