Search found 12 matches

by mphuget
Thu May 07, 2020 10:00 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Generating all the moves on a board
Replies: 2
Views: 885

Generating all the moves on a board

Hello everyone, I am doing in this moment move predictions with neural network and for the labels, that is, the move, I decided to opt for one-hot encoding (which seems to be the best choice). Some authors consider using all the labels we got from the training and doing the one-hot encoding on these...
by mphuget
Fri Apr 10, 2020 6:56 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Count the number of nodes of perft(14) and beyond
Replies: 17
Views: 3106

Re: Count the number of nodes of perft(14) and beyond

Hello, Dann (and all), I guess it could be worthwhile to decide which details we need and then I change my code to obtain them My first requirement would be to fill the table as shown on https://www.chessprogramming.org/Perft_Results. I hate empty spaces :-) This exercise has four interests for me: ...
by mphuget
Sun Apr 05, 2020 11:00 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Count the number of nodes of perft(14) and beyond
Replies: 17
Views: 3106

Re: Count the number of nodes of perft(14) and beyond

Hello Dann,

I know Dann but I want to try Perft using GPU as did Ankan by the past, first using his approach and try to run it on the cloud, then provide my own approach

Cheers,
mph
by mphuget
Sat Apr 04, 2020 9:31 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Count the number of nodes of perft(14) and beyond
Replies: 17
Views: 3106

Count the number of nodes of perft(14) and beyond

Hello everyone, By curiosity (or not) I am wondering how it is possible to store the number of nodes when we count perft(14) and beyond: the number of nodes is greater than 2^64, do we consider an overflow and we let the count continue and we add on 128 bits when it is finished? I remember Steven Ed...
by mphuget
Fri Jan 10, 2020 8:09 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Trying to understand Retrograde analysis (EGTB generation)
Replies: 10
Views: 1805

Re: Trying to understand Retrograde analysis (EGTB generation)

Thanks, Dann for your answer. Actually, I don't want to propose another format, Syzygy works fine for me and the reason for this post is curiosity. When I will manage to play backward, I could have a better idea of how multithreading could be used to improve the speed, even if I am sure https://gith...
by mphuget
Fri Jan 10, 2020 7:08 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Trying to understand Retrograde analysis (EGTB generation)
Replies: 10
Views: 1805

Trying to understand Retrograde analysis (EGTB generation)

Hello all, I try to figure out retrograde analysis (especially when generating EGTB). From all the mate positions, going backward on the first move is easy, who checks the king, and I generate all possible moves from this attacking piece to every square on the board leading to mate in 1. But when th...
by mphuget
Tue Dec 10, 2019 2:17 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: I bought a stop-gap machine
Replies: 34
Views: 4495

Re: I bought a stop-gap machine

Well, I am about to retire Ezekiel (my 64 core box that does about 47M NPS on the Asmfish bench). This new box will do more than twice the NPS at 1/3 of the power. If you want to pay for crating and shipping, I would send you that machine (but keep the sound proof box). Warning, it weighs 100KG, so...
by mphuget
Fri Nov 29, 2019 9:11 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Pawn move generation in bitboards
Replies: 3
Views: 1354

Pawn move generation in bitboards

Hello all, I try to modify my engine to use Bitboards (currently 10x12 board representation) and I am stuck in understanding what is the most efficient for pawn move generation. I have a U64 representation for all the pieces on the board I have a U64 pawn representation for one side For every bit se...
by mphuget
Fri Oct 25, 2019 8:31 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Another take on DTS?
Replies: 20
Views: 4483

Another take on DTS?

Hello everyone, I am surprised to see currently so few interests in Dynamic Tree Splitting in comparison with Lazy SMP. Is it because DTS seems to be adapted to multiprocessors or large-scale network while Lazy SMP is for multithreading? Is there any scientific or empirical comparison between the tw...