Search found 605 matches

by petero2
Mon Jun 29, 2020 5:21 pm
Forum: Computer Chess Club: General Topics
Topic: Website engines
Replies: 24
Views: 1042

Re: Website engines

Hi Ed,

The official Texel repository is actually here:

https://github.com/peterosterlund2/texel

The link you have is from someone else who uploaded source code for released Texel versions (nothing wrong with that) before I made the full Texel history available on github.
by petero2
Fri May 08, 2020 9:49 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Crafty 25.6 search stability
Replies: 39
Views: 5744

Re: Crafty 25.6 search stability

Do ALL threads sit at the "same table" when accessing (say) the transposition/refutation table (hash table)? Yes, there is only "one table", in order to guarantee deterministic results. If so this would really seem to slow things down. Suppose one of them is in the q-search and doesn't even want to...
by petero2
Tue May 05, 2020 4:59 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Crafty 25.6 search stability
Replies: 39
Views: 5744

Re: Crafty 25.6 search stability

I don't see how that can solve the problem. You say "always access the TT in the same order". Take two threads, A and B. If they access at the same time, you can certainly make it A then B. But what if B does a probe and A has not gotten there yet. In fact, you don't even know that A will EVER get ...
by petero2
Sun May 03, 2020 9:36 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Crafty 25.6 search stability
Replies: 39
Views: 5744

Re: Crafty 25.6 search stability

The threads are not synchronized, if they were performance would be terrible. I agree, but a little less terrible than what you suggest here: About all you can do is allow each thread to search one move and then pause while the other threads do the same, one at a time. And in the same order. Which ...
by petero2
Sun Mar 08, 2020 9:45 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Strange sporadic speed limitation in engine running in Linux on Ryzen
Replies: 19
Views: 3144

Re: Strange sporadic speed limitation in engine running in Linux on Ryzen

I recommend to see the following video about what can affect program performance: https://www.youtube.com/watch?v=r-TLSBdHe1A Many things can affect the memory layout of the program and affect its performance, including the directory where it runs from. In particular for your situation: Does the co...
by petero2
Sun Mar 01, 2020 12:44 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Measuring Hash Collisions (once again)
Replies: 37
Views: 5110

Re: Measuring Hash Collisions (once again)

That pretty much confirms that 16 (non-redundant) bits in the signature is already good enough from an Elo perspective. And with full verification of the hash move, Elo would be the only concern. Some of the bits in the Zobrist key are used to determine which bucket a position belongs to. Each entr...
by petero2
Sat Feb 29, 2020 12:22 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Measuring Hash Collisions (once again)
Replies: 37
Views: 5110

Re: Measuring Hash Collisions (once again)

But the question is: "what is good enough"? I would say that 1 collision in 64M nodes is already far better than you would need to have totally undetectable impact on Elo. According to Bob 1 collison in 10K nodes should already have no measurable impact, so it seems you are several thousand times b...
by petero2
Thu Feb 27, 2020 8:21 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Don't understand CarryRippler
Replies: 14
Views: 1900

Re: Don't understand CarryRippler

Yes, i understand now. I know a number wraps around at the boundary: u8 255 + 1 = 0 u8 0 - 1 = 255 I didn't realize the carry rippler expects this, and that C implicitly allows it when using unsigned variables. Thanks for the help :) This is actually explicitly allowed by the C language specificati...
by petero2
Tue Feb 18, 2020 7:36 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: hash collisions
Replies: 370
Views: 31564

Re: hash collisions

Oh. In that case the chess equivalent might be Sting, an engine based on Stockfish 2.1.1 where all known Stockfish 2.1.1 bugs would have been fixed so any bug that it'd have would need to have survived since that version. So a bug-free version might be possible, but at what cost? (in this case, no ...
by petero2
Tue Feb 18, 2020 7:15 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: hash collisions
Replies: 370
Views: 31564

Re: hash collisions

Ovyron wrote:
Tue Feb 18, 2020 6:21 pm
Any other example of "bug-free software"
TeX version 3.14159265 might come pretty close. The latest bug fix release is 6 years old and the bounty for finding a bug is $327.68.