| View previous topic :: View next topic |
| Author |
Message |
Daniel Shawul
Joined: 14 Mar 2006 Posts: 2187 Location: Ethiopia
|
Post subject: Re: Zeta CL uses Magic Bitboards Posted: Tue Jun 12, 2012 1:46 pm |
|
|
Maybe he could have taken the legality test out side the inner loop since it is better to do pseudo-legal move generation first. There is the majic multiplier and then piece extraction from the quad bb which all seem born out of need and can't do anything about.
| Quote: |
For storing some intermediate results inside the local scope of a routine, say 32 bitboards as local variables is a bottleneck on a GPU?
|
Number of registers used are the bottleneck. The nvcc compiler does so much optimization with the registers it is very hard to free up a single register. Temporaries are quickly reused and manual optimization for me did not work, still stuck at 45 registers or so for chess i.e including move generation,make and everything else. Hex is far lower about 12 register IIRC so it is the nature of the game IMO. If you do a completely "disjoint" set of operations that don't share registers in the same kernel, the register usage will not increase. For example I have a move generator and number of moves counter, which almost does the same calculations but doesn't save the move. The two consume about 30 registers each but my register usage still remains the same. So I would be curious to see how much can be saved by taking out the legality testing outside if any.
Anyway Kogge-stone is good iff you are going to make the fills in the 8 directions using a core each. As Srdja mentioned he tested that and the focus is on using each core to do its own move generation and search. Well with the the trend I saw from the new device Kepler, maybe that approach of unrolling loops and working with kogge stone will be better. Too many number of cores, too little memory, lower clock rates and addition of dynamic parallelism that unrolls loops automatically and stuff like that. My favourite for general computing is still the fermi ... _________________ https://sites.google.com/site/dshawul/
https://github.com/dshawul |
|
| Back to top |
|
 |
|
| Subject |
Author |
Date/Time |
Is there such a thing as branchless move generation? |
John Hamlen |
Thu Jun 07, 2012 7:50 pm |
Re: Is there such a thing as branchless move generation? |
Gerd Isenberg |
Thu Jun 07, 2012 8:44 pm |
Re: Is there such a thing as branchless move generation? |
John Hamlen |
Thu Jun 07, 2012 10:18 pm |
Re: Is there such a thing as branchless move generation? |
Edmund Moshammer |
Thu Jun 07, 2012 10:57 pm |
Re: Is there such a thing as branchless move generation? |
John Hamlen |
Fri Jun 08, 2012 8:21 am |
Re: Is there such a thing as branchless move generation? |
Sven Schüle |
Fri Jun 08, 2012 9:21 am |
Re: Is there such a thing as branchless move generation? |
John Hamlen |
Fri Jun 08, 2012 10:31 am |
Re: Is there such a thing as branchless move generation? |
Edmund Moshammer |
Fri Jun 08, 2012 10:41 am |
Re: Is there such a thing as branchless move generation? |
John Hamlen |
Sat Jun 09, 2012 1:17 pm |
Re: Is there such a thing as branchless move generation? |
Daniel Shawul |
Fri Jun 08, 2012 11:12 am |
Re: Is there such a thing as branchless move generation? |
Edmund Moshammer |
Fri Jun 08, 2012 11:18 am |
Re: Is there such a thing as branchless move generation? |
Daniel Shawul |
Fri Jun 08, 2012 3:39 pm |
Re: Is there such a thing as branchless move generation? |
Vincent Diepeveen |
Sat Jun 16, 2012 9:07 am |
Re: Is there such a thing as branchless move generation? |
Daniel Shawul |
Sat Jun 16, 2012 11:27 am |
Re: Is there such a thing as branchless move generation? |
Vincent Diepeveen |
Sat Jun 16, 2012 11:35 am |
Re: Is there such a thing as branchless move generation? |
John Hamlen |
Sat Jun 09, 2012 2:43 pm |
Re: Is there such a thing as branchless move generation? |
Daniel Shawul |
Sat Jun 09, 2012 6:02 pm |
Re: Is there such a thing as branchless move generation? |
John Hamlen |
Sun Jun 10, 2012 2:21 pm |
Re: Is there such a thing as branchless move generation? |
Daniel Shawul |
Sun Jun 10, 2012 6:33 pm |
Re: Is there such a thing as branchless move generation? |
Daniel Shawul |
Sun Jun 10, 2012 7:16 pm |
Re: Is there such a thing as branchless move generation? |
Edmund Moshammer |
Sun Jun 10, 2012 9:24 pm |
Re: Is there such a thing as branchless move generation? |
Daniel Shawul |
Sun Jun 10, 2012 9:54 pm |
Re: Is there such a thing as branchless move generation? |
John Hamlen |
Sun Jun 10, 2012 10:14 pm |
Re: Is there such a thing as branchless move generation? |
Daniel Shawul |
Mon Jun 11, 2012 12:49 am |
Re: Is there such a thing as branchless move generation? |
John Hamlen |
Mon Jun 11, 2012 12:43 pm |
Re: Is there such a thing as branchless move generation? |
Daniel Shawul |
Mon Jun 11, 2012 2:41 pm |
Re: Is there such a thing as branchless move generation? |
John Hamlen |
Mon Jun 11, 2012 10:12 pm |
Re: Is there such a thing as branchless move generation? |
Daniel Shawul |
Mon Jun 11, 2012 11:12 pm |
Re: Is there such a thing as branchless move generation? |
John Hamlen |
Mon Jun 11, 2012 11:41 pm |
Re: Is there such a thing as branchless move generation? |
Vincent Diepeveen |
Sat Jun 16, 2012 9:56 am |
Re: Is there such a thing as branchless move generation? |
John Hamlen |
Sun Jun 10, 2012 10:58 pm |
Re: Is there such a thing as branchless move generation? |
Gerd Isenberg |
Sat Jun 09, 2012 7:48 pm |
Re: Is there such a thing as branchless move generation? |
John Hamlen |
Sun Jun 10, 2012 5:28 pm |
Re: Is there such a thing as branchless move generation? |
Edmund Moshammer |
Fri Jun 08, 2012 10:32 am |
Re: Is there such a thing as branchless move generation? |
John Hamlen |
Sat Jun 09, 2012 2:57 pm |
Re: Is there such a thing as branchless move generation? |
Vincent Diepeveen |
Sat Jun 16, 2012 8:57 am |
Re: Zeta CL uses Magic Bitboards |
Srdja Matovic |
Mon Jun 11, 2012 2:53 pm |
Re: Zeta CL uses Magic Bitboards |
Gerd Isenberg |
Mon Jun 11, 2012 8:55 pm |
Re: Zeta CL uses Magic Bitboards |
Srdja Matovic |
Mon Jun 11, 2012 9:35 pm |
Re: Zeta CL uses Magic Bitboards |
Daniel Shawul |
Mon Jun 11, 2012 11:26 pm |
Re: Zeta CL uses Magic Bitboards |
Srdja Matovic |
Tue Jun 12, 2012 1:16 am |
Re: Zeta CL uses Magic Bitboards |
Gerd Isenberg |
Tue Jun 12, 2012 6:58 am |
Re: Zeta CL uses Magic Bitboards |
Daniel Shawul |
Tue Jun 12, 2012 1:46 pm |
Re: Zeta CL uses Magic Bitboards |
Srdja Matovic |
Tue Jun 12, 2012 2:35 pm |
Re: Zeta CL uses Magic Bitboards |
Daniel Shawul |
Tue Jun 12, 2012 4:37 pm |
Re: Zeta CL uses Magic Bitboards |
Vincent Diepeveen |
Sat Jun 16, 2012 8:46 am |
Re: Zeta CL uses Magic Bitboards |
Daniel Shawul |
Sat Jun 16, 2012 11:17 am |
Re: Zeta CL uses Magic Bitboards |
Vincent Diepeveen |
Sat Jun 16, 2012 11:30 am |
Re: Zeta CL uses Magic Bitboards |
Daniel Shawul |
Sat Jun 16, 2012 1:56 pm |
Re: Zeta CL uses Magic Bitboards |
Vincent Diepeveen |
Sat Jun 16, 2012 2:51 pm |
Re: Zeta CL uses Magic Bitboards vs Kogge-Stone |
Srdja Matovic |
Tue Jun 12, 2012 1:05 am |
Re: Zeta CL uses Magic Bitboards vs Kogge-Stone |
Gerd Isenberg |
Tue Jun 12, 2012 6:31 am |
Re: Zeta CL uses Magic Bitboards |
John Hamlen |
Mon Jun 11, 2012 10:35 pm |
Re: Zeta CL uses Magic Bitboards |
Daniel Shawul |
Mon Jun 11, 2012 11:00 pm |
Re: Is there such a thing as branchless move generation? |
Vincent Diepeveen |
Sat Jun 16, 2012 8:40 am |
Re: Is there such a thing as branchless move generation? |
Vincent Diepeveen |
Sat Jun 16, 2012 9:23 am |
|
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
|
|