Author: Omid David
Date: 19:07:38 10/03/02
Go up one level in this thread
On October 03, 2002 at 21:30:56, Jeremiah Penery wrote: >On October 03, 2002 at 20:39:02, Omid David wrote: > >>On October 03, 2002 at 20:29:28, Jeremiah Penery wrote: >> >>>On October 03, 2002 at 19:39:27, Omid David wrote: >>> >>>>BTW, how much (in percentage or NPS) did the implemented draw detection slow >>>>down the engine? > >>> The only difference in the two benches below is the blocked-position draw >>>detection. > >>>With detection: >>>Total nodes: 69410117 >>>Raw nodes per second: 708266 >>>Total elapsed time: 98 >>> >>>Without: >>>Total nodes: 69410117 >>>Raw nodes per second: 715568 >>>Total elapsed time: 97 > >>Quite impressive, just 1.02% slower (paractically negligible). Do you invoke >>that detection at every node in the search tree? > >Yes, but most of the time it exits very quickly (a couple of 'if' clauses). I >could probably make it cost even less, but I'm not very interested in doing that >yet. But the problem may arise in the endgames where there are a number of pawns and pieces left which can potentially cause a "block draw". Then in the draw detection heuristic, almost in every node you'll pass the first few 'if's and will have to step into a detailed assessment of the position. How much does it slow down there, in comparison to the non-detecting version?
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.