Search found 190 matches

by Sesse
Tue Jan 28, 2020 4:15 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Why is MultiPV so slow?
Replies: 22
Views: 714

Re: Why is MultiPV so slow?

Heh, so it depends on the position?

I wonder; when we change positions, do we automatically get the 8 ply penalty for hash table replacement? Could turning that off help spv search win more of the time?
by Sesse
Mon Jan 27, 2020 11:41 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Why is MultiPV so slow?
Replies: 22
Views: 714

Re: Why is MultiPV so slow?

For future reference: I tried integrating manual multi-PV in my scripts, and it was markedly slower than just regular multi-PV. I have no idea why it didn't work this time; perhaps it was the different position, perhaps these things pan out very differently with many threads, perhaps the hash size m...
by Sesse
Sun Jan 26, 2020 9:32 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Why is MultiPV so slow?
Replies: 22
Views: 714

Re: Why is MultiPV so slow?

Note that I have a different need for MultiPV from what a typical chess player would; I want to communicate to the viewer (on analysis.sesse.net) the score for every possible move. Typical use cases are “why can't he just capture that piece?” and “is this a situation where the player needs to find a...
by Sesse
Sat Jan 25, 2020 7:05 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Why is MultiPV so slow?
Replies: 22
Views: 714

Re: Why is MultiPV so slow?

Later moves tend to be worse. Worse, means more fail-lows. More fail-lows means more time spent to resolve them. So it's not unusual to have the later moves being more costly to search and absorbing a disproportionate part of the "nodes budget". Hm. So a problem is that the engine spends a dispropo...
by Sesse
Sat Jan 25, 2020 5:59 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Why is MultiPV so slow?
Replies: 22
Views: 714

Re: Why is MultiPV so slow?

That's interesting! But shouldn't it still be sub-linear, since the later moves should be cheaper to search (better move ordering, fewer moves to search) than the main PV?
by Sesse
Sat Jan 25, 2020 5:44 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Why is MultiPV so slow?
Replies: 22
Views: 714

Re: Why is MultiPV so slow?

OK, so potentially we're either looking at something wrong in my timing, or a bona fide Stockfish bug?
by Sesse
Sat Jan 25, 2020 4:27 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Why is MultiPV so slow?
Replies: 22
Views: 714

Re: Why is MultiPV so slow?

Yes, the point of setting MultiPV to 100 is to get all moves, so it's intentional. I'm a bit confused at the importance of it being inside iterative deepening, though; I assumed iterative deepening was only run at the root?
by Sesse
Sat Jan 25, 2020 2:28 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Why is MultiPV so slow?
Replies: 22
Views: 714

Why is MultiPV so slow?

So, of course I get why MultiPV is much slower than a regular single-PV search; you can't use the cutoffs and have to research suboptimal moves to score them. But still, the slowdown seems extreme. Right now, I'm searching the same position on Stockfish 11 on very similar hardware (basically two-soc...
by Sesse
Sat Dec 14, 2019 7:14 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: An AMD compiling hunch
Replies: 12
Views: 1123

Re: An AMD compiling hunch

Are we also going to call instructions like XLAT "phony baloney" also because it perform poorly (even on Intel?) XLAT is pretty pony baloney! It's a relic of the 8086 days that wasn't really a good idea even back then. The idea that GCC doesnt have a say is wrong. The goal of the compiler should be...
by Sesse
Tue Dec 10, 2019 2:34 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: An AMD compiling hunch
Replies: 12
Views: 1123

Re: An AMD compiling hunch

It isn't really GCC's fault. Stockfish simply goes and asks the compiler (well, preprocessor) “does the CPU I'm compiling for support BMI2?”, and if so, sends BMI2 code to the compiler. GCC has no say in the matter once it has given a truthful answer. (Not that there's a good way to ask “does the CP...