Search found 257 matches

by Cardoso
Sun Jul 07, 2019 3:51 pm
Forum: Computer Chess Club: General Topics
Topic: Ryzen 3000 series is out
Replies: 30
Views: 2715

Ryzen 3000 series is out

Hi,
AMD 3000 series cpus look good for AB engines.
We have no chess benchmarks yet but things do look good.
https://www.youtube.com/watch?v=z3aEv3EzMyQ

best regards,
Alvaro
by Cardoso
Mon Apr 22, 2019 1:18 pm
Forum: Computer Chess Club: General Topics
Topic: SSD Problem
Replies: 7
Views: 2553

Re: SSD Problem

That's a very good SSD. It sure is not a problem with the NAND itself. It looks like changing OS broke things. It shouldn't happen, but technology can fail spectaculary. https://www.easeus.com/data-recovery/other-recovery-software/SSD-drive-recovery.htm https://www.techradar.com/best/best-free-data-...
by Cardoso
Thu Apr 18, 2019 5:59 pm
Forum: Computer Chess Club: General Topics
Topic: SSD Problem
Replies: 7
Views: 2553

Re: SSD Problem

First of all could you please say the brand and model, as well as the rated max TBW? Not all brands have the same resilience, for example I have red ADATA SSDs failed well before the "expected end of life". I think it was not a problem with the NAND itself but with other components. Also how long ha...
by Cardoso
Mon Feb 04, 2019 11:33 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Killers and forward pruning test searches
Replies: 16
Views: 2921

Re: Killers and forward pruning test searches

Hmm, Topple loses about 10 Elo if I implement this. It uses less nodes for the same depth on some positions, but more nodes on other positions. Strange. RubiChess also loses Elo when clearing killers of ply+1. Clearing at ply+2 is more promising but tests still running. ./Andreas I guess this is ve...
by Cardoso
Mon Feb 04, 2019 5:34 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Killers and forward pruning test searches
Replies: 16
Views: 2921

Re: Killers and forward pruning test searches

konsolas wrote:
Mon Feb 04, 2019 4:41 pm
Hmm, Topple loses about 10 Elo if I implement this. It uses less nodes for the same depth on some positions, but more nodes on other positions. Strange.
Could you please post the number of games you used and the statistics details?
by Cardoso
Sun Feb 03, 2019 4:43 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Killers and forward pruning test searches
Replies: 16
Views: 2921

Re: Killers and forward pruning test searches

Killers are very local. Some engines even enforce that (presumably because it was shown to gain Elo) by explicitly clearing the killer slots for the daughter level when they enter a node. This means you could never inherit a killer from any node other than a sibbling in the same search. Then the si...
by Cardoso
Sun Jan 27, 2019 11:33 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Killers and forward pruning test searches
Replies: 16
Views: 2921

Killers and forward pruning test searches

When we do a reduced/test search to try to make a beta cuttof like Null Move Search / ProbCut / Multi-Cut Pruning, Internal iterative deepening (for getting a best move to try first), etc, it came to my mind that killers generated inside those searches must be of inferior quality because of the redu...
by Cardoso
Sat Aug 25, 2018 8:35 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: singular extension
Replies: 9
Views: 1705

Re: singular extension

I must say I think I tried both and do not see a great elo gain... As stockfish comments "+60elo" inside the source code in front of singular extension, I was hopping for something better. It would be very interesting making a new test using SF9 on singular extensions playing 20k+ games. Perhaps on...
by Cardoso
Sun Jun 10, 2018 10:32 pm
Forum: Computer Chess Club: General Topics
Topic: Worlds most powerful supercomputer.
Replies: 9
Views: 1840

Re: Worlds most powerful supercomputer.

Very nice, for years I've been reading news about colossal machines with astronomical computing power that cost millions of dollars to built and a big amount also to maintain and run. Now, how about putting this machine to really good use like finding the cure for cancer? If mankind knows so much ab...
by Cardoso
Thu May 31, 2018 2:18 pm
Forum: Computer Chess Club: Programming and Technical Discussions
Topic: Move ordering based on stored values of child nodes
Replies: 4
Views: 1202

Re: Move ordering based on stored values of child nodes

With current move ordering heuristics, you get a cut-off from the first three moves in 95-99% of cases. That's the statistic you have to beat. Do you get that consistently throughout the entire game? Do you get that for example from the starting position or after the first say 10 moves? And what ab...