Search found 2052 matches

by mar
Fri Apr 10, 2020 6:48 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: if statement and calculation faster than constant
Replies: 43
Views: 2293

Re: if statement and calculation faster than constant

My experience is that you cannot really optimize programs this way (i.e. make very small changes and see how they affect execution time) on these modern multi-scalar out-of-order CPUs. Execution times of instructions just depend too much on each other. E.g. on the order in which instructions appear...
by mar
Thu Apr 09, 2020 12:05 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Rewriting RomiChess from scratch for SMP
Replies: 17
Views: 1818

Re: Rewriting RomiChess from scratch for SMP

That's easy to solve by only considering recaptures after QS depth X. It also prevents going through pointless queen plunder raids in normal game positions. I even generate non-capture evasions in QS if a capture gives check, but limit that to one ply less than the recapture-only mode, and don't do...
by mar
Thu Apr 09, 2020 11:04 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Rewriting RomiChess from scratch for SMP
Replies: 17
Views: 1818

Re: Rewriting RomiChess from scratch for SMP

(except for Bob in Crafty at root for move ordering - and I think even this is a bad idea) At root, I'm doing that also, but apart from the question whether it helps, at least it doesn't take noticeable time. the problem I have with this is that when you're searching a position with qs explosion (s...
by mar
Thu Apr 09, 2020 10:31 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Rewriting RomiChess from scratch for SMP
Replies: 17
Views: 1818

Re: Rewriting RomiChess from scratch for SMP

Ras wrote:
Thu Apr 09, 2020 9:31 am
Are you sure that Demolito really calls QS on every move right after generation?
why would anybody want do this? (except for Bob in Crafty at root for move ordering - and I think even this is a bad idea)
by mar
Thu Apr 09, 2020 5:40 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Rewriting RomiChess from scratch for SMP
Replies: 17
Views: 1818

Re: Rewriting RomiChess from scratch for SMP

This seems efficient as this has to be done on all leaf nodes regardless. This is horribly inefficient because if you get a fail-high on the first move of a leaf node, then you have done the computation for all the other moves in vain. I'd go with a pseudo legal move generator and do the legality c...
by mar
Tue Mar 31, 2020 2:15 am
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Hash entry/bucket memory usage optimization
Replies: 7
Views: 795

Re: Hash entry/bucket memory usage optimization

Can I be certain that this way of verifying that I got the correct entry within the bucket holds up on higher depths? If so, I can generalize this has table (leaf_nodes -> data, etc)... and use the same code when doing the search. In that case, my hash table is already done before I even finish the...
by mar
Fri Mar 27, 2020 11:50 am
Forum: Computer Chess Club: General Topics
Topic: New Tool
Replies: 55
Views: 6981

Re: New Tool

Update: mea doesn't run under wine, so it won't run under Linux but mrl.exe runs fine. mea.exe seems to be a program compiled into a python binary it seems it turns out that mea is actually ferdy's open source https://github.com/fsmosca/Multiple-move-Epd-Analyzer and Ed's tool is probably built on t...
by mar
Fri Mar 27, 2020 10:41 am
Forum: Computer Chess Club: General Topics
Topic: New Tool
Replies: 55
Views: 6981

Re: New Tool

it's not a passion, but a religion (biased, naturally)
make silly demands, spit in every direction because there is only one true "god" :D sounds familiar?

(yodeling, really? that's interesting :)
by mar
Fri Mar 27, 2020 9:59 am
Forum: Computer Chess Club: General Topics
Topic: New Tool
Replies: 55
Views: 6981

Re: New Tool

flok wrote:
Fri Mar 27, 2020 9:53 am
I'd rather not use closed-source software coming from a windows platform.
Then don't use it, the world doesn't revolve around Linux fanboys :)
(to be clear, I like Linux, but I hate Linux fanboys)