Author: Reinhard Scharnagl
Date: 07:35:01 12/04/05
Go up one level in this thread
On December 04, 2005 at 10:29:21, Reinhard Scharnagl wrote: >On December 04, 2005 at 09:18:53, Brian Richardson wrote: > >>On December 03, 2005 at 22:21:20, Reinhard Scharnagl wrote: >> >>>On December 03, 2005 at 22:14:20, Brian Richardson wrote: >>> >>>>How similar are perft and alpha/beta trees? >>>>Can perft caching can really be used to guide PVS or other actual search >>>>caching? >>> >>>The test was targeting to find a relation between intended thinking time and >>>cache size. Thus the absence of specialized search strategies has been intended. >>> >>>Reinhard. >> >>I thought perft just counts moves to a specified depth, so I'm not clear on what >>"intended thinking time" means in this context. Of course, this could just be a >>minor ENFLI English not first language issue :) >>[As in someone who knows 2 languages is bilingual, and 3 is trilingual, but only >>1 language is an American] >> >>Clearly, it is fair to say that caching does significantly impact perft time to >>depth performance (although I suspect few engines use cache for perft). If you like I could switch to use German language, making less mistakes. German is the most spoken native language in pure European countries. >a) see: http://www.talkchess.com/forums/2/message.html?197740 > >b) how do you test, that your transposition table is working well and correct? >My method is to compare Perft results to detect errors e.g. in generating of >hash keys or in the replacement scheme's overall performance. c) here is an example doing much more than a genuine Perft of such a comparing: FEN: rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1 +-*--b--c--d--*--f--g--*-+ MS Vis.Studio C++ Vers. 13.10 8 |[r][n][b][q][k][b][n][r]| (Compilation: Dec 4 2005) 7 |[p][p][p][p][p][p][p][p]| 6 | ::: ::: ::: :::| Perft Testseries 5 |::: ::: ::: ::: | (With TT Caching 256.0 MB / 4-fold) 4 | ::: ::: ::: :::| TT Access Success 48.5% 3 |::: ::: ::: ::: | 2 |<P><P><P><P><P><P><P><P>| Smirf Test No.: 00 1 |<R><N><B><Q><K><B><N><R>| =>+-*--b--c--d--*--f--g--*-+ Break Time 133.0 Sec. Ply Nodes all (x) (e.p.) all (+) (#) Prom. Castl. Sec. ------------------------------------------------------------------------------ 1 20 0 0 0 0 0 0 0 2 400 0 0 0 0 0 0 0 3 8902 34 0 12 0 0 0 0 4 197281 1576 0 469 8 0 0 0 5 4865609 82719 258 27351 347 0 0 0.1 6 119060324 2812008 5248 809099 10828 0 0 1.4 7 3195901860 108329926 319617 33103848 435767 0 883453 17.0 8 84998978956 3523740106 7187977 968981593 9852036 0 23605205 232.5 ------------------------------------------------------------------------------ FEN: rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1 +-*--b--c--d--*--f--g--*-+ MS Vis.Studio C++ Vers. 13.10 8 |[r][n][b][q][k][b][n][r]| (Compilation: Dec 3 2005) 7 |[p][p][p][p][p][p][p][p]| 6 | ::: ::: ::: :::| Perft Testseries 5 |::: ::: ::: ::: | 4 | ::: ::: ::: :::| (without caching) 3 |::: ::: ::: ::: | 2 |<P><P><P><P><P><P><P><P>| Smirf Test No.: 00 1 |<R><N><B><Q><K><B><N><R>| =>+-*--b--c--d--*--f--g--*-+ Break Time 133.0 Sec. Ply Nodes all (x) (e.p.) all (+) (#) Prom. Castl. Sec. ------------------------------------------------------------------------------ 1 20 0 0 0 0 0 0 0 2 400 0 0 0 0 0 0 0 3 8902 34 0 12 0 0 0 0 4 197281 1576 0 469 8 0 0 0 5 4865609 82719 258 27351 347 0 0 0.2 6 119060324 2812008 5248 809099 10828 0 0 4.6 7 3195901860 108329926 319617 33103848 435767 0 883453 124.0 8 84998978956 3523740106 7187977 968981593 9852036 0 23605205 3242.0 ------------------------------------------------------------------------------ Reinhard.
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.