TalkChess.com

According to FabianVDW (of FabChess, which is also an engine in Rust), mutex locking works, but is too slow if an engine uses more than 8 threads. Then the overhead of the locks/unlocks becomes so big that more threads don't add any more speed. He says that "racey access to the hash table is necess...

Ethereal, Weiss, and now Fabchess have found large elo gains tuning using these datasets. Weiss was able to implement the tuner a la the paper and found +50 elo in 60s+.6s with the initial tune. I believe Weiss was already tuning using an older dataset of mine, and with his own python framework. An...

I can chime in on two of the points and maybe provide a bit of a mathematical background, although I am unsure how much of this you know already. 1) If I want the NN to return an evaluation, which could be used in a traditional alphabeta search engine, how is this achieved most commonly? Do I have a...

Thanks for this tourney Graham! Fab overperformed pretty hard, but still almost managed to throw the lead away in the end

Very interesting, thank you for the background information bob. Although, I think it is similar to telling your child that they should / shouldn't do something. Invariably they ignore good advice and have to learn the hard way rather than by their parents experience :D - I am sure I will look back ...

According to your number it looks like the relative speed of perft with/without bulk counting depends on one's implementation. I definitely second this! If you do extra work on initializing extra info on your GameState struct when constructing it for instance, it won't hurt search performance at al...

While correctly working code is most important, 14M nps seems a little on the slow side. Tinker is far from the fastest at perft, but it is about 95M nps on a 3.6GHz CPU. Move generation in Tinker is done with magic bitboards, BTW. 95M nps without bulk counting is very fast. 14 Mnps looks like a go...

FabChess has a hard time on this one, needs 14 minutes to get it right on 4 threads info depth 1 seldepth 0 nodes 0 nps 0 hashfull 0 time 0 score cp 31 pv d5c4 info depth 2 seldepth 0 nodes 0 nps 0 hashfull 0 time 0 score cp 74 pv d5c4 g3g4 info depth 3 seldepth 0 nodes 0 nps 0 hashfull 0 time 0 sco...

Thanks for testing
I will try to resolve the errors for the next version

Thanks for having tested FabChess Very good results