Search found 641 matches

by smatovic
Mon Jul 04, 2011 7:01 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: SPPS - a Simple Parallel Processing Scheme
Replies: 7
Views: 1828

Re: GPUs and unsynched SIMD Units

fyi: i managed to get a non-recursive, negamax-like search algorithm running on one SIMD Unit of a GPU. But GPUs consist of tens of these SIMD Units and a effective sync between them is simply not possible. So the idea to pack the hole computation on the GPU is a stopper, somehow the CPU has to be i...
by smatovic
Fri Jul 01, 2011 7:17 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: magic in fruit
Replies: 14
Views: 3232

Re: magic in fruit

if you are new to chessprogramming i suggest to start with an easier to implement move generator and board presentation like mailbox or 0x88.

if you get mailbox running than you can extend your engine with bitboards.

--
Srdja
by smatovic
Tue Jun 28, 2011 3:12 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: SPPS - a Simple Parallel Processing Scheme
Replies: 7
Views: 1828

Re: DPPS - a Dynamic Parallel Processing Scheme

Here a simplified version of the DPPS i use. // for every search depth while (sd >= 0) { // and for each move/board with a finite search depth while (movecounter[sd] > 0 && sd < max_depth) { decrease(movecounter[sd]); // get an board for Process boardindex = getBoardInd...
by smatovic
Fri Jun 24, 2011 10:39 am
Forum: Computer Chess Club: General Topics
Topic: Zeta plays chess on a gpu
Replies: 10
Views: 3162

Re: Zeta plays chess on a gpu

Hi Srdja, I would like to know a little more about your engine. Some time ago I corresonded to a company who was very interested in gpu's and chess and at that time was willing to allow the use of some high end equipment all using gpus. Back then their interest was to generate 7 piece egtb's. If I ...
by smatovic
Fri Jun 24, 2011 10:30 am
Forum: Computer Chess Club: General Topics
Topic: Zeta plays chess on a gpu
Replies: 10
Views: 3162

Re: Zeta plays chess on a gpu

Does it work with any video card? Or just with nVidias? (perhaps you use CUDA?) I use OpenCL, so It works on GPUs. CPUs, APUs... :-) every Device with an "OpenCL" Implementation, sufficient RAM and able to process min 128 threads. AMD for example has excellent OpenCL Implementation for CPUs. -- Srdja
by smatovic
Fri Jun 24, 2011 10:23 am
Forum: Computer Chess Club: General Topics
Topic: Zeta plays chess on a gpu
Replies: 10
Views: 3162

Re: Zeta plays chess on a gpu

very nice to read that you have managed to create a chess playing engine on a GPU! :-) What is the reason for this? Could you explain how your "SPPS" scheme is related to Q-Search? Since you do not want to split a QS node over different threads there must be some other reason which I do not underst...
by smatovic
Thu Jun 23, 2011 4:43 pm
Forum: Computer Chess Club: General Topics
Topic: Zeta plays chess on a gpu
Replies: 10
Views: 3162

Re: Zeta plays chess on a gpu

Did you submit your engine to AMD OpenCL coding competition innovation challenge: http://community.topcoder.com/amdapp/am ... etition-2/ ?
Not yet.
Thank you for the hint :-)

--
Srdja
by smatovic
Thu Jun 23, 2011 11:43 am
Forum: Computer Chess Club: General Topics
Topic: Zeta plays chess on a gpu
Replies: 10
Views: 3162

Re: Zeta plays chess on a gpu

How many nodes per second do you get with 128 threads? With its current "SPPS" search, which is comparable to a Negamax wo AlphaBeta pruning Zeta achieves with 128 Threads ~ 500.000 nodes per second. No Quiscence Search, no Castling/En Passant moves. Very simple Eval. Did you calculate the speedup ...
by smatovic
Thu Jun 23, 2011 11:22 am
Forum: Computer Chess Club: General Topics
Topic: Zeta plays chess on a gpu
Replies: 10
Views: 3162

Zeta plays chess on a gpu

Juhu,

finally i got the negamax scores runnning and now my little engine plays chess on a gpu :)

http://zeta-chess.blogspot.com/

Still a lot of work to do...therefore i am looking for sponsors, hardware, money, what ever.

--
Srdja
by smatovic
Thu Jun 23, 2011 12:54 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: SPPS - a Simple Parallel Processing Scheme
Replies: 7
Views: 1828

Re: SPPS - a Simple Parallel Processing Scheme

pseudo code update with negamax-like scores, it looks somekind of ugly compared to recursion, but for those who are interested it should give an idea how to use it on a simd-device... // pid => Process ID, 0-127 globalboards[sd*128]; globalscores[sd*128]; globalmoves[sd*128*128]; // for each search ...