| View previous topic :: View next topic |
| Author |
Message |
Sven Schüle
Joined: 15 May 2008 Posts: 2246 Location: Berlin, Germany
|
Post subject: Re: C# Performance Posted: Sun Feb 05, 2012 12:51 am |
|
|
| RoadWarrior wrote: |
| Sven Schüle wrote: |
Just to be sure: did this change as I described it above cause 2% slowdown, or 2% speedup? I could not imagine how avoiding to test E8 (resp. E1) for attacks twice can slow down your engine. If it does indeed, then how did you implement it? |
I added IsInCheck before generating any moves. That appears to have cost more time than it saved in avoiding the generation of castling moves. Mind you, I haven't profiled it yet, but those are the only 2 changes I made. |
Well, it is clear that calling IsInCheck at every node costs much more than avoiding castling generation, since there are several reasons why castling can be not available that are cheaper to test than testing attacks to the king. But that is not a fair comparison ...
My proposal was meant as:
a) IF you already have the IsInCheck prior to move generation then there is no need to test for attacks to E8 (E1) in castling move generation, just test the inCheck flag, and
b) if you don't have IsInCheck and you don't want to use that information in the move generator then you can still save time by not testing attacks to E8 (E1) twice, for two castle types per color. You might need an additional flag like "attacksToE8WereTested" in the castling generation function to achieve that.
Since in fact you did not have IsInCheck before, the only change where you can compare speed is b).
For perft you do not strictly need IsInCheck, depending on your legality checking strategy of course. But for real search you will very likely need it anyway. I see perft as a tool to validate (or debug) the "real" move generator that will be part of your searching engine, so I would build the move generator exactly in the way it is meant to work in the real engine (including information that was collected in advance), and then base "perft" on that. The "perft" validation shall still work the same way when you have your search up and running, in my opinion the target is not to make "perft" as fast as it can be, or you lose the ability to test any later changes to board representation, make/unmake, or move generator by running "perft".
Sven |
|
| Back to top |
|
 |
|
| Subject |
Author |
Date/Time |
C# Performance |
Richard Allbert |
Fri Jan 27, 2012 8:41 pm |
Re: C# Performance |
Kevin Hearn |
Fri Jan 27, 2012 9:01 pm |
Re: C# Performance |
Richard Allbert |
Fri Jan 27, 2012 9:13 pm |
Re: C# Performance |
Gary |
Fri Jan 27, 2012 9:07 pm |
Re: C# Performance |
Richard Allbert |
Fri Jan 27, 2012 9:17 pm |
Re: C# Performance |
Gary |
Fri Jan 27, 2012 10:09 pm |
Re: C# Performance |
Mark Pearce |
Sat Jan 28, 2012 12:59 am |
Re: C# Performance |
Richard Allbert |
Sat Jan 28, 2012 8:44 am |
Re: C# Performance |
Sven Schüle |
Sat Jan 28, 2012 9:48 am |
Re: C# Performance |
Richard Allbert |
Sat Jan 28, 2012 10:25 am |
Re: C# Performance |
Richard Allbert |
Sat Jan 28, 2012 11:41 am |
Re: C# Performance |
Mark Pearce |
Sat Jan 28, 2012 11:44 am |
Re: C# Performance |
Richard Allbert |
Sat Jan 28, 2012 10:31 am |
Re: C# Performance |
Marco Costalba |
Sat Jan 28, 2012 1:09 pm |
Re: C# Performance |
Lucas Braesch |
Sat Jan 28, 2012 1:52 pm |
Re: C# Performance |
Richard Allbert |
Sat Jan 28, 2012 2:05 pm |
Re: C# Performance |
Marco Costalba |
Sat Jan 28, 2012 2:30 pm |
Re: C# Performance |
Richard Allbert |
Sat Jan 28, 2012 5:22 pm |
Re: C# Performance |
Sven Schüle |
Sat Jan 28, 2012 6:02 pm |
Re: C# Performance |
Richard Allbert |
Sat Jan 28, 2012 6:12 pm |
Re: C# Performance |
Robert Purves |
Sat Jan 28, 2012 10:17 pm |
Re: C# Performance |
Marco Costalba |
Sat Jan 28, 2012 10:25 pm |
Re: C# Performance |
Thomas Petzke |
Sat Jan 28, 2012 4:21 pm |
Re: C# Performance |
Richard Allbert |
Mon Jan 30, 2012 12:52 pm |
Re: C# Performance |
Marco Costalba |
Mon Jan 30, 2012 6:37 pm |
Re: C# Performance |
Richard Allbert |
Wed Feb 01, 2012 6:55 pm |
Re: C# Performance |
Marco Costalba |
Wed Feb 01, 2012 7:18 pm |
Re: C# Performance |
Mark Pearce |
Mon Jan 30, 2012 9:55 pm |
Re: C# Performance |
Richard Allbert |
Wed Feb 01, 2012 6:57 pm |
Re: C# Performance |
Mark Pearce |
Wed Feb 01, 2012 9:59 pm |
Re: C# Performance |
Mark Pearce |
Fri Feb 03, 2012 1:34 am |
Re: C# Performance |
Robert Purves |
Fri Feb 03, 2012 3:26 am |
Re: C# Performance |
Mark Pearce |
Fri Feb 03, 2012 9:28 am |
Re: C# Performance |
Sven Schüle |
Fri Feb 03, 2012 9:45 am |
Re: C# Performance |
Mark Pearce |
Sat Feb 04, 2012 7:57 pm |
Re: C# Performance |
Sven Schüle |
Sat Feb 04, 2012 10:36 pm |
Re: C# Performance |
Mark Pearce |
Sat Feb 04, 2012 11:49 pm |
Re: C# Performance |
Sven Schüle |
Sun Feb 05, 2012 12:51 am |
Re: C# Performance |
Mark Pearce |
Sun Feb 05, 2012 1:05 am |
Re: C# Performance |
Sven Schüle |
Fri Feb 03, 2012 10:04 am |
Re: C# Performance |
Mark Pearce |
Fri Feb 03, 2012 11:40 am |
Re: C# Performance |
Richard Allbert |
Sat Feb 04, 2012 7:10 pm |
Re: C# Performance |
Lucas Braesch |
Sat Feb 04, 2012 8:31 pm |
Re: C# Performance |
Mark Pearce |
Sat Feb 04, 2012 10:23 pm |
Re: C# Performance |
Mark Pearce |
Sun Feb 05, 2012 12:23 am |
Re: C# Performance |
Lucas Braesch |
Sun Feb 05, 2012 2:57 am |
Re: C# Performance |
Mark Pearce |
Sun Feb 05, 2012 11:40 am |
Re: C# Performance |
Richard Allbert |
Mon Feb 06, 2012 6:29 pm |
Re: C# Performance |
Richard Allbert |
Mon Feb 06, 2012 6:29 pm |
Re: C# Performance |
Richard Allbert |
Mon Feb 06, 2012 6:30 pm |
Re: C# Performance |
Richard Allbert |
Mon Feb 06, 2012 6:45 pm |
Re: C# Performance |
Gary |
Mon Feb 06, 2012 7:00 pm |
Re: C# Performance |
Richard Allbert |
Mon Feb 06, 2012 7:12 pm |
Re: C# Performance |
Sven Schüle |
Mon Feb 06, 2012 10:50 pm |
Re: C# Performance |
Richard Allbert |
Tue Feb 07, 2012 9:52 am |
Re: C# Performance |
Richard Allbert |
Tue Feb 07, 2012 9:53 am |
Re: C# Performance |
Richard Allbert |
Tue Feb 07, 2012 9:53 am |
Re: C# Performance |
Richard Allbert |
Tue Feb 07, 2012 9:54 am |
Re: C# Performance |
Mark Pearce |
Tue Feb 07, 2012 1:59 pm |
Re: C# Performance |
Richard Allbert |
Tue Feb 07, 2012 2:54 pm |
Re: C# Performance |
Richard Allbert |
Tue Feb 07, 2012 3:06 pm |
Re: C# Performance |
Mark Pearce |
Tue Feb 07, 2012 5:17 pm |
Re: C# Performance |
Richard Allbert |
Tue Feb 07, 2012 6:10 pm |
Re: C# Performance |
Richard Allbert |
Tue Feb 07, 2012 6:20 pm |
Re: C# Performance |
Mark Pearce |
Tue Feb 07, 2012 8:02 pm |
Re: C# Performance |
Richard Allbert |
Tue Feb 07, 2012 9:19 pm |
Re: C# Performance |
Mark Pearce |
Tue Feb 07, 2012 10:20 pm |
Re: C# Performance |
Richard Allbert |
Wed Feb 08, 2012 7:47 am |
Re: C# Performance |
Richard Allbert |
Wed Feb 08, 2012 7:57 am |
Re: C# Performance |
Sven Schüle |
Wed Feb 08, 2012 10:09 am |
Re: C# Performance |
Mark Pearce |
Wed Feb 08, 2012 12:16 pm |
Re: C# Performance |
Vincent Diepeveen |
Sat Feb 18, 2012 10:03 pm |
Re: C# Performance |
Mark Pearce |
Wed Feb 08, 2012 9:30 am |
Re: C# Performance |
Richard Allbert |
Wed Feb 08, 2012 12:02 pm |
Re: C# Performance |
Richard Allbert |
Wed Feb 08, 2012 12:34 pm |
Re: C# Performance |
Mark Pearce |
Wed Feb 08, 2012 12:56 pm |
Re: C# Performance |
Richard Allbert |
Wed Feb 08, 2012 1:18 pm |
Re: C# Performance |
Richard Allbert |
Wed Feb 08, 2012 2:30 pm |
Re: C# Performance |
Richard Allbert |
Wed Feb 08, 2012 2:34 pm |
Re: C# Performance |
Mark Pearce |
Wed Feb 08, 2012 3:02 pm |
Re: C# Performance |
Richard Allbert |
Wed Feb 08, 2012 3:27 pm |
Re: C# Performance |
Richard Allbert |
Wed Feb 08, 2012 3:29 pm |
Re: C# Performance |
Richard Allbert |
Wed Feb 08, 2012 3:59 pm |
Re: C# Performance |
Richard Allbert |
Wed Feb 08, 2012 4:04 pm |
Re: C# Performance |
Mark Pearce |
Wed Feb 08, 2012 6:07 pm |
Re: C# Performance |
Richard Allbert |
Thu Feb 09, 2012 10:41 am |
Re: C# Performance |
Mark Pearce |
Thu Feb 09, 2012 1:47 pm |
Re: C# Performance |
Richard Allbert |
Thu Feb 09, 2012 2:56 pm |
Re: C# Performance |
Mark Pearce |
Sat Feb 11, 2012 4:41 pm |
Re: C# Performance |
Richard Allbert |
Sun Feb 12, 2012 8:34 am |
Re: C# Performance |
Lucas Braesch |
Sun Feb 12, 2012 10:52 am |
Re: C# Performance |
Richard Allbert |
Sun Feb 12, 2012 12:26 pm |
Re: C# Performance |
Ron Murawski |
Mon Feb 13, 2012 5:26 am |
Re: C# Performance |
Lucas Braesch |
Mon Feb 13, 2012 5:52 am |
Re: C# Performance |
Richard Allbert |
Mon Feb 13, 2012 6:27 am |
Re: C# Performance |
Ron Murawski |
Tue Feb 14, 2012 6:38 am |
Re: C# Performance |
Lucas Braesch |
Tue Feb 14, 2012 8:22 am |
Re: C# Performance |
Mark Pearce |
Tue Feb 14, 2012 11:30 pm |
Re: C# Performance |
Richard Allbert |
Thu Feb 16, 2012 10:49 am |
Re: C# Performance |
Mark Pearce |
Sat Feb 18, 2012 12:34 pm |
Re: C# Performance |
Richard Allbert |
Mon Feb 20, 2012 9:19 pm |
Re: C# Performance |
Mark Pearce |
Tue Feb 21, 2012 8:33 am |
Re: C# Performance |
Richard Allbert |
Tue Feb 21, 2012 6:37 pm |
Re: C# Performance |
Vincent Diepeveen |
Sat Feb 18, 2012 10:19 pm |
Re: C# Performance |
Mark Pearce |
Sun Feb 12, 2012 2:01 pm |
Re: C# Performance |
Richard Allbert |
Sun Feb 12, 2012 4:40 pm |
Re: C# Performance |
Vincent Diepeveen |
Sat Feb 18, 2012 10:06 pm |
Re: C# Performance |
Mark Pearce |
Sun Feb 19, 2012 1:01 pm |
Re: C# Performance |
Vincent Diepeveen |
Sun Feb 19, 2012 5:25 pm |
Re: C# Performance |
Mark Pearce |
Sun Feb 19, 2012 7:04 pm |
Re: C# Performance |
Vincent Diepeveen |
Sun Feb 19, 2012 9:17 pm |
Re: C# Performance |
Mark Pearce |
Sun Feb 19, 2012 11:33 pm |
Re: C# Performance |
Richard Allbert |
Sun Feb 19, 2012 9:22 pm |
Re: C# Performance |
Vincent Diepeveen |
Sun Feb 19, 2012 9:32 pm |
Re: C# Performance |
Vincent Diepeveen |
Sun Feb 19, 2012 9:37 pm |
Re: C# Performance |
Vincent Diepeveen |
Sun Feb 19, 2012 10:06 pm |
Re: C# Performance |
Lucas Braesch |
Wed Feb 08, 2012 10:56 am |
Re: C# Performance |
Mark Pearce |
Wed Feb 08, 2012 11:30 am |
Re: C# Performance |
Richard Allbert |
Wed Feb 08, 2012 12:04 pm |
Re: C# Performance |
Tony Soares |
Tue Feb 07, 2012 3:47 am |
Re: C# Performance |
Richard Allbert |
Tue Feb 07, 2012 9:55 am |
Re: C# Performance |
Mark Pearce |
Sat Feb 04, 2012 8:37 pm |
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
|