Stockfish version with hash saving capability

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Dann Corbit, Harvey Williamson

Forum rules
This textbox is used to restore diagrams posted with the [d] tag before the upgrade.
User avatar
cdani
Posts: 2190
Joined: Sat Jan 18, 2014 9:24 am
Location: Andorra
Contact:

Re: Stockfish version with hash saving capability

Post by cdani » Sun Aug 20, 2017 10:53 pm

cdani wrote:Here you have an example of an epd file:
http://talkchess.com/forum/viewtopic.ph ... 76&t=64914
This version of Stockfish only uses:
fen
acd (depth searched)
bm (best move)
ce (score)

Rodolfo Leoni
Posts: 545
Joined: Tue Jun 06, 2017 2:49 pm
Location: Italy

Re: Stockfish version with hash saving capability

Post by Rodolfo Leoni » Mon Aug 21, 2017 11:07 am

cdani wrote:
cdani wrote:Here you have an example of an epd file:
http://talkchess.com/forum/viewtopic.ph ... 76&t=64914
This version of Stockfish only uses:
fen
acd (depth searched)
bm (best move)
ce (score)
Thanks! I'll test it this evening. :)

BTW, the idea of hash merging isn't so esoteric. Stockfish PA GTB has a hash merge command, and it was possible because the PH file isn't usually a huge file. It looks a sequential file, so I guess it does nothing else than appending positions and their data to file, without trying to detect any duplicate. With your SF savehash system a user would need a lot of RAM for such a procedure (and Peter's hardware has it).

I'll try another experiment too this evening. I'll set hash size to 2 GBs, then I'll load two different hashes of 1 GB size each, to check if both were loaded or if second one would overwrite first one.

Later. :)
F.S.I. Chess Teacher

User avatar
cdani
Posts: 2190
Joined: Sat Jan 18, 2014 9:24 am
Location: Andorra
Contact:

Re: Stockfish version with hash saving capability

Post by cdani » Mon Aug 21, 2017 12:05 pm

Rodolfo Leoni wrote: BTW, the idea of hash merging isn't so esoteric. Stockfish PA GTB has a hash merge command, and it was possible because the PH file isn't usually a huge file. It looks a sequential file, so I guess it does nothing else than appending positions and their data to file, without trying to detect any duplicate. With your SF savehash system a user would need a lot of RAM for such a procedure (and Peter's hardware has it).
What happens is that the hash that this Stockfish saves is the main hash, and is absolutely unrelated to the one saved by PA GTB. Is not possible to merge it, as the content is partial.
Rodolfo Leoni wrote: I'll try another experiment too this evening. I'll set hash size to 2 GBs, then I'll load two different hashes of 1 GB size each, to check if both were loaded or if second one would overwrite first one.
Will be overwritten, of course, and the main hash will be resized to 1 GB. The way it works is not related at all to the one of PA GTB. Is not like a text file that one can append, replace...

Rodolfo Leoni
Posts: 545
Joined: Tue Jun 06, 2017 2:49 pm
Location: Italy

Re: Stockfish version with hash saving capability

Post by Rodolfo Leoni » Mon Aug 21, 2017 12:11 pm

cdani wrote:
Rodolfo Leoni wrote: BTW, the idea of hash merging isn't so esoteric. Stockfish PA GTB has a hash merge command, and it was possible because the PH file isn't usually a huge file. It looks a sequential file, so I guess it does nothing else than appending positions and their data to file, without trying to detect any duplicate. With your SF savehash system a user would need a lot of RAM for such a procedure (and Peter's hardware has it).
What happens is that the hash that this Stockfish saves is the main hash, and is absolutely unrelated to the one saved by PA GTB. Is not possible to merge it, as the content is partial.
Rodolfo Leoni wrote: I'll try another experiment too this evening. I'll set hash size to 2 GBs, then I'll load two different hashes of 1 GB size each, to check if both were loaded or if second one would overwrite first one.
Will be overwritten, of course, and the main hash will be resized to 1 GB. The way it works is not related at all to the one of PA GTB. Is not like a text file that one can append, replace...
It's clear. I already suspected it was that way but I sometimes learn by having experiences. :)
F.S.I. Chess Teacher

zullil
Posts: 6442
Joined: Mon Jan 08, 2007 11:31 pm
Location: PA USA
Full name: Louis Zulli

Re: Stockfish version with hash saving capability

Post by zullil » Mon Aug 21, 2017 12:14 pm

duncan wrote:
zullil wrote:
duncan wrote:
zullil wrote: asmFish eventually found mate-in-11 from the original position. But it took a while.

Code: Select all

info depth 54 seldepth 24 multipv 1 time 5491130 nps 50956798 score mate 11 nodes 279810404462 hashfull 999 tbhits 0 pv a2a4 h7h5 d2d4 c7c6 e2e4 g7g6 d4d5 g6g5 d1h5 e8d8 h5f7 c6d5 f1b5 d5e4 f7d5 d8c8 d5d7 c8b8 b5a6 b8a8 d7b7

is your nps 50956798 which is very fast and quicker than threadripper.? May I ask what sytem you are using?
Not sure what your first sentence is trying to ask. :wink: Yes, the nodes/sec is 50956798, which is indeed very fast. I don't know anything about Threadripper. I have 2 x Intel(R) Xeon(R) CPU E5-2687W v3 @ 3.10GHz and 64GB RAM. I tend to use asmFish because it is NUMA-aware, which seems to improve performance on my system.
the following position is a mater in 33 and is the Longest mate in King, Bishop and Knight versus King endgame.

it takes me a long time on my computer to find the mate in 33. how long do you estimate it should take on your computer .just 10 minutes ?

[d]K7/2kB4/8/8/8/8/8/5N2 w - - 0 1

http://www.gilith.com/chess/endgames/kbn_k.html
The time until mate-in-33 is reported varies quite a bit, due to the randomness inherent in multithreaded searches. The shortest time was about 1.5 minutes and the longest was about 5 minutes. The following seems typical:

Code: Select all

info depth 64 seldepth 72 multipv 1 time 129908 nps 73529464 score mate 33 nodes 9552065660 hashfull 67 tbhits 0 pv d7h3 c7b6 h3g2 b6c5 a8b7 c5d6 b7b6 d6e6 b6c6 e6e5 g2d5 e5f4 c6d6 f4f5 f1e3 f5f4 e3c4 f4g5 d6e7 g5g6 d5e4 g6g7 c4e5 g7g8 e7f6 g8h8 e4d5 h8h7 e5g6 h7h6 d5g8 h6h5 g6e5 h5h4 g8d5 h4g3 e5d3 g3g4 f6g6 g4h4 d5f3 h4g3 f3d1 g3g2 g6g5 g2g3 d1e2 g3h3 g5f4 h3h4 d3e1 h4h3 e2g4 h3h2 f4f3 h2h1 g4h3 h1h2 f3g4 h2g1 g4g3 g1h1 h3g2 h1g1 e1f3
This was done using a 4 GB hash table. The size of the hash will also affect the solution time.

Vinvin
Posts: 4919
Joined: Thu Mar 09, 2006 8:40 am
Full name: Vincent Lejeune

Re: Stockfish version with hash saving capability

Post by Vinvin » Mon Aug 21, 2017 1:24 pm

duncan wrote:the following position is a mater in 33 and is the Longest mate in King, Bishop and Knight versus King endgame.

it takes me a long time on my computer to find the mate in 33. how long do you estimate it should take on your computer .just 10 minutes ?

[d]K7/2kB4/8/8/8/8/8/5N2 w - - 0 1

http://www.gilith.com/chess/endgames/kbn_k.html
Sting-SF 8.5, 4*4GHz. Mate in 60 in 4m17s, mate in 33 in 14m39s.

Code: Select all

Sting-sf-85:
...
 39	00:39	 591.018k	15.005k	+77,22	1.Fh3 Rd6 2.Rb7 Rc5 3.Ce3 Rd4 4.Cc2+ Rc3 5.Ce1 Rb2 6.Cd3+ Ra2 7.Rc6 Rb1 8.Ce5 Ra1 9.Cc4 Ra2 10.Ff5 Rb3 11.Rd5 Rc3 12.Fe4 Rb4 13.Rd4 Rb3 14.Cd6 Ra4 15.Rc5 Ra3 16.Cc4+ Ra2 17.Rd5 Rb3 18.Ff5
 40	02:05	 1.879.217k	14.956k	+77,27	1.Fh3 Rd6 2.Fc8 Re5 3.Ce3 Re4 4.Cc2 Rd3 5.Ff5+ Rd2 6.Fe4 Rd1 7.Rb7 Rd2 8.Ra7 Rc1 9.Rb8 Rb1 10.Rb7 Rb2 11.Rc7 Rb1 12.Rb6 Rc1 13.Rc6 Rd1 14.Rb7
 41	02:22	 2.140.361k	14.999k	+77,27	1.Fh3 Rd6 2.Fc8 Re5 3.Ce3 Re4 4.Cc2 Rd3 5.Ff5+ Rd2 6.Fe4 Rd1 7.Rb7 Rd2 8.Ra7 Rc1 9.Rb8 Rb1 10.Rb7 Rb2 11.Rb6 Rb1 12.Rc5 Rb2 13.Rc4 Ra2 14.Cd4 Ra3 15.Fg2 Rb2 16.Rb4 Rc1 17.Rc3 Rb1 18.Cc2 Ra2 19.Fd5+ Rb1 20.Fc4 Rc1 21.Fd5 Rd1 22.Cd4 Re1 23.Fe4 Rd1 24.Fd3 Re1 25.Cf5 Rf2 26.Rd4 Rf3 27.Rd5 Rf4 28.Fe4 Rg5 29.Re5 Rg4 30.Fd5 Rg5 31.Fc6 Rg6 32.Fg2 Rf7 33.Fh3 Rg6 34.Cd4 Rh7 35.Rd5 Rg7 36.Rc6 Rg8 37.Rd7 Rf7 38.Cf3 Rf6 39.Rc7 Rf7 40.Rd8 Rg6 41.Re7 Rh6 42.Fe6 Rh7 43.Fc8 Rh6 44.Rd6 Rg6 45.Re6 Rg7 46.Re7 Rg6 47.Fd7 Rg7 48.Cd2 Rg6 49.Ce4 Rg7 50.Fe6 Rg6
 42	02:37	 2.363.798k	15.034k	+77,27	1.Fh3 Rd6 2.Fc8 Re5 3.Ce3 Re4 4.Cc2 Rd3 5.Ff5+ Rd2 6.Fe4 Rd1 7.Rb7 Rd2 8.Ra7 Rc1 9.Rb8 Rb1 10.Rb7 Rb2 11.Rb6 Rb3 12.Rc5 Rc3 13.Rd5 Rb3 14.Rd4 Ra2 15.Rc3 Rb1 16.Fd5 Rc1 17.Cb4 Rb1 18.Fe6 Ra1 19.Fc4 Rb1 20.Cd3 Ra1 21.Cf2 Rb1 22.Fd5 Rc1 23.Cd3+ Rb1 24.Ce5 Rc1 25.Fe4 Rd1 26.Cg4 Re2 27.Rd4 Rd1 28.Ch2 Re1 29.Ff5 Re2 30.Fd7 Rf2 31.Fe6 Rg3 32.Cf1+ Rg2 33.Cd2 Rh1 34.Rc5 Rh2 35.Ff5 Rg3 36.Rd4 Rf2 37.Ce4+ Rg2 38.Fe6 Rf3 39.Fd7 Rg2 40.Ff5 Rh2 41.Fe6 Rg2 42.Cd2
 43	04:17	 3.994.020k	15.517k	+M60	1.Fh3 Rd6 2.Rb7 Rc5 3.Rc7 Rd5 4.Rd7 Re5 5.Cd2 Rd4 6.Cb1 Rd3 7.Ff5+ Rc4 8.Fe6+ Rd4 9.Fg8 Re5 10.Re7 Rd4 11.Rf6 Re4 12.Fb3 Rd4 13.Re6 Re3 14.Re5 Rf3 15.Fc2 Re3 16.Fh7 Rf3 17.Fd3 Rg3 18.Cd2 Rg4 19.Cc4 Rh5 20.Ff5 Rh6 21.Rf6 Rh5 22.Ce3 Rh6 23.Cf1 Rh5 24.Re5 Rh4 25.Ce3 Rh5 26.Cf1
 44	04:20	 4.047.070k	15.508k	+M47	1.Fh3 Rd6 2.Rb7 Rc5 3.Rc7 Rd5 4.Cd2 Rc5 5.Fe6 Rd4 6.Rd6 Rd3 7.Cb1 Rc2 8.Ff5+ Rb3 9.Rd5 Rb2 10.Rc4 Ra1 11.Cd2 Rb2 12.Cb3 Ra3 13.Rc3 Ra2 14.Fe4 Ra3 15.Fb1 Ra4 16.Cd4 Ra5 17.Rc4 Rb6 18.Cb5 Rb7 19.Rc5 Rc8 20.Rd6 Rb7 21.Fe4+ Ra6 22.Rc5 Ra5 23.Ff5 Ra4 24.Fc2+ Ra5 25.Cd6 Ra6 26.Fe4 Ra7 27.Cb5+ Ra6 28.Ca3 Ra7 29.Cc2 Ra6 30.Ce3 Ra7 31.Cc4 Rb8 32.Rd6 Rc8 33.Ce5 Rd8 34.Fc6 Rc8 35.Cd7 Rd8 36.Cc5 Rc8 37.Fd7+ Rb8 38.Fe6 Ra7 39.Rc6 Rb8 40.Rb6 Ra8 41.Ff5
 45	04:22	 4.064.312k	15.503k	+M46	1.Fh3 Rd6 2.Rb7 Rc5 3.Rc7 Rd5 4.Cd2 Rc5 5.Fe6 Rd4 6.Rd6 Rd3 7.Cb1 Rc2 8.Ff5+ Rb3 9.Rd5 Rb2 10.Rc4 Ra1 11.Cd2 Rb2 12.Cb3 Ra3 13.Rc3 Ra2 14.Fe4 Ra3 15.Fb1 Ra4 16.Cd4 Ra5 17.Rc4 Rb6 18.Cb5 Rb7 19.Rc5 Rc8 20.Rd6 Rb7 21.Fe4+ Ra6 22.Rc5 Ra5 23.Fc6 Ra4 24.Fd5 Ra5 25.Ca3 Ra4 26.Cb1 Ra5 27.Cc3 Ra6 28.Fe4 Ra7 29.Fc6 Rb8 30.Rd6 Ra7 31.Cd5 Ra6 32.Ce3 Ra5 33.Rc5 Ra6 34.Cc4 Ra7 35.Rd6 Rb8 36.Rd7 Ra7 37.Rc7 Ra6 38.Fd7 Ra7 39.Fc8 Ra8 40.Cd6
 46	04:24	 4.101.555k	15.495k	+M43	1.Fh3 Rd6 2.Rb7 Rc5 3.Rc7 Rd5 4.Cd2 Rc5 5.Fe6 Rd4 6.Rd6 Rd3 7.Cb1 Rc2 8.Ff5+ Rb3 9.Rd5 Rb2 10.Rd4 Rb3 11.Fe4 Rb2 12.Cd2 Ra2 13.Rc3 Ra1 14.Fd3 Ra2 15.Cb3 Ra3 16.Fb1 Ra4 17.Cd4 Ra5 18.Rc4 Rb6 19.Cb5 Rb7 20.Ff5 Rc6 21.Fh3 Rb6 22.Fg2 Ra5 23.Fd5 Rb6 24.Rb4 Ra6 25.Rc5 Ra5 26.Ca3 Ra4 27.Cb1 Ra5 28.Cc3 Ra6 29.Fe4 Ra7 30.Fc6 Rb8 31.Rd6 Ra7 32.Cd5 Ra6 33.Ce3 Ra5 34.Rc5 Ra6 35.Cc4 Ra7 36.Rd6 Rb8 37.Rd7 Ra7 38.Rc7 Ra6 39.Fd7 Ra7 40.Fc8 Ra8 41.Cd6
 47	04:29	 4.165.226k	15.478k	+M43	1.Fh3 Rd6 2.Rb7 Rc5 3.Rc7 Rd5 4.Cd2 Rc5 5.Fe6 Rd4 6.Rd6 Rd3 7.Cb1 Rc2 8.Ff5+ Rb3 9.Rd5 Rb2 10.Rd4 Rb3 11.Fe4 Rb2 12.Cd2 Ra2 13.Rc3 Ra1 14.Fd3 Ra2 15.Cb3 Ra3 16.Fb1 Ra4 17.Cd4 Ra5 18.Rc4 Rb6 19.Cb5 Rb7 20.Ff5 Rc6 21.Fh3 Rb6 22.Fg2 Ra5 23.Fd5 Rb6 24.Rb4 Ra6 25.Rc5 Ra5 26.Ca3 Ra4 27.Cb1 Ra5 28.Cc3 Ra6 29.Fe4 Ra7 30.Fc6 Rb8 31.Rd6 Ra7 32.Cd5 Ra6 33.Ce3 Ra5 34.Rc5 Ra6 35.Cc4 Ra7 36.Rd6 Rb8 37.Rd7 Ra7 38.Rc7 Ra6 39.Fd7 Ra7 40.Fc8 Ra8 41.Cd6
 48	04:33	 4.223.100k	15.425k	+M43	1.Fh3 Rd6 2.Rb7 Rc5 3.Rc7 Rd5 4.Cd2 Rc5 5.Fe6 Rd4 6.Rd6 Rd3 7.Cb1 Rc2 8.Ff5+ Rb3 9.Rd5 Rb2 10.Rd4 Rb3 11.Fe4 Rb2 12.Cd2 Ra2 13.Rc3 Ra1 14.Fd3 Ra2 15.Cb3 Ra3 16.Fb1 Ra4 17.Cd4 Ra5 18.Rc4 Rb6 19.Cb5 Rb7 20.Ff5 Rc6 21.Fh3 Rb6 22.Fg2 Ra5 23.Fd5 Rb6 24.Rb4 Ra6 25.Rc5 Ra5 26.Ca3 Ra4 27.Cb1 Ra5 28.Cc3 Ra6 29.Fe4 Ra7 30.Fc6 Rb8 31.Rd6 Ra7 32.Cd5 Ra6 33.Ce3 Ra5 34.Rc5 Ra6 35.Cc4 Ra7 36.Rd6 Rb8 37.Rd7 Ra7 38.Rc7 Ra6 39.Fd7 Ra7 40.Fc8 Ra8 41.Cd6
 49	06:18	 5.720.933k	15.111k	+M38	1.Fh3 Rc6 2.Rb8 Rb6 3.Ce3 Rc6 4.Cc2 Rd6 5.Rb7 Rd5 6.Rc7 Re5 7.Rd7 Rf6 8.Rd6 Rg7 9.Fe6 Rh6 10.Cd4 Rg6 11.Re5 Rg7 12.Cf5+ Rg6 13.Fd5 Rh7 14.Rf6 Rh8 15.Ch4 Rh7 16.Cg6 Rh6 17.Fg8 Rh5 18.Ce5 Rh4 19.Fd5 Rg3 20.Cd3 Rg4 21.Rg6 Rh4 22.Fb3 Rg3 23.Fd1 Rh3 24.Rf5 Rg3 25.Re5 Rh2 26.Rf4 Rg1 27.Fg4 Rg2 28.Ff5 Rg1 29.Rf3 Rf1 30.Fh3+ Rg1 31.Cf4 Rh1 32.Rf2 Rh2 33.Fg4 Rh1 34.Ch5
 50	06:22	 5.771.378k	15.093k	+M37	1.Fh3 Rc6 2.Rb8 Rb6 3.Ce3 Rc6 4.Rc8 Rd6 5.Rd8 Re5 6.Re7 Rd4 7.Cc2+ Re4 8.Rd6 Rd3 9.Ff5+ Rc4 10.Re5 Rc3 11.Rd5 Rd2 12.Rd4 Re2 13.Fd3+ Rf3 14.Re5 Rg4 15.Fe4 Rh4 16.Ff5 Rh5 17.Ce3 Rh6 18.Rf6 Rh5 19.Fe4 Rh4 20.Rf5 Rh5 21.Cg4 Rh4 22.Rf4 Rh5 23.Ff5 Rh4 24.Fc2 Rh5 25.Fe4 Rh4 26.Cf6 Rh3 27.Ch5 Rh2 28.Cg3 Rh3 29.Cf5
 51	06:28	 5.854.071k	15.073k	+M37	1.Fh3 Rc6 2.Rb8 Rb6 3.Ce3 Rc6 4.Rc8 Rd6 5.Rd8 Re5 6.Re7 Rd4 7.Cc2+ Re4 8.Rd6 Rd3 9.Ff5+ Rc4 10.Re5 Rc3 11.Rd5 Rd2 12.Rd4 Re2 13.Fd3+ Rf3 14.Re5 Rg4 15.Fe4 Rh4 16.Ff5 Rh5 17.Ce3 Rh6 18.Rf6 Rh5 19.Fe4 Rh4 20.Rf5 Rh5 21.Cg4 Rh4 22.Rf4 Rh5 23.Ff5 Rh4 24.Fc2 Rh5 25.Fe4 Rh4 26.Cf6 Rh3 27.Ch5 Rh2 28.Cg3 Rh3 29.Cf5
 52	06:36	 5.967.158k	15.044k	+M37	1.Fh3 Rc6 2.Rb8 Rb6 3.Ce3 Rc6 4.Rc8 Rd6 5.Rd8 Re5 6.Re7 Rd4 7.Cc2+ Re4 8.Rd6 Rd3 9.Ff5+ Rc4 10.Re5 Rc3 11.Rd5 Rb3 12.Rd4 Rb2 13.Rc4 Rb1 14.Rc3 Ra2 15.Fe4 Rb1 16.Fd5 Rc1 17.Fa2 Rd1 18.Cd4 Re1 19.Rd3 Rf2 20.Ce2 Rg2 21.Fe6 Rf3 22.Ff5 Rf2 23.Fg4 Re1 24.Cg3 Rf2 25.Cf5 Re1 26.Re3 Rf1 27.Ff3 Rg1 28.Ch4 Rf1 29.Fe2+ Rg1 30.Rf3 Rh1 31.Rf2 Rh2 32.Cf5 Rh3 33.Fh5 Rh2 34.Fg4 Rh1 35.Cg3+ Rh2 36.Cf1+ Rh1 37.Ff3+
 53	06:45	 6.086.867k	15.029k	+M37	1.Fh3 Rc6 2.Rb8 Rb6 3.Ce3 Rc6 4.Rc8 Rd6 5.Rd8 Re5 6.Re7 Rd4 7.Cc2+ Re4 8.Rd6 Rd3 9.Ff5+ Rc4 10.Re5 Rc3 11.Rd5 Rb3 12.Rd4 Rb2 13.Rc4 Rb1 14.Rc3 Ra2 15.Fe4 Rb1 16.Fd5 Rc1 17.Fa2 Rd1 18.Cd4 Re1 19.Rd3 Rf2 20.Ce2 Rg2 21.Fe6 Rf3 22.Ff5 Rf2 23.Fg4 Re1 24.Cg3 Rf2 25.Cf5 Re1 26.Re3 Rf1 27.Ff3 Rg1 28.Ch4 Rf1 29.Fe2+ Rg1 30.Rf3 Rh1 31.Rf2 Rh2 32.Cf5 Rh3 33.Fh5 Rh2 34.Fg4 Rh1 35.Cg3+ Rh2 36.Cf1+ Rh1 37.Ff3+
 54	07:06	 6.385.565k	14.971k	+M37	1.Fh3 Rb6 2.Rb8 Rc6 3.Ce3 Rb6 4.Rc8 Rc5 5.Rd8 Rb4 6.Cc2+ Rc5 7.Re7 Rc4 8.Ff5 Rc3 9.Re6 Rb2 10.Re5 Rc3 11.Rd5 Rb3 12.Rd4 Rb2 13.Rc4 Rb1 14.Rc3 Ra2 15.Fe4 Rb1 16.Fd5 Rc1 17.Fa2 Rd1 18.Cd4 Re1 19.Rd3 Rf2 20.Ce2 Rg2 21.Fe6 Rf3 22.Ff5 Rf2 23.Fg4 Re1 24.Cg3 Rf2 25.Cf5 Re1 26.Re3 Rf1 27.Ff3 Rg1 28.Ch4 Rf1 29.Fe2+ Rg1 30.Rf3 Rh1 31.Rf2 Rh2 32.Cf5 Rh3 33.Fh5 Rh2 34.Fg4 Rh1 35.Cg3+ Rh2 36.Cf1+ Rh1 37.Ff3+
 55	07:19	 6.550.660k	14.907k	+M37	1.Fh3 Rb6 2.Rb8 Rc6 3.Ce3 Rb6 4.Rc8 Rc5 5.Rd8 Rb4 6.Cc2+ Rc5 7.Re7 Rc4 8.Ff5 Rc3 9.Re6 Rb2 10.Re5 Rc3 11.Rd5 Rb3 12.Rd4 Rb2 13.Rc4 Rb1 14.Rc3 Ra2 15.Fe4 Rb1 16.Fd5 Rc1 17.Fa2 Rd1 18.Cd4 Re1 19.Rd3 Rf2 20.Ce2 Rg2 21.Fe6 Rf3 22.Ff5 Rf2 23.Fg4 Re1 24.Cg3 Rf2 25.Cf5 Re1 26.Re3 Rf1 27.Ff3 Rg1 28.Ch4 Rf1 29.Fe2+ Rg1 30.Rf3 Rh1 31.Rf2 Rh2 32.Cf5 Rh3 33.Fh5 Rh2 34.Fg4 Rh1 35.Cg3+ Rh2 36.Cf1+ Rh1 37.Ff3+
 56	07:36	 6.787.956k	14.863k	+M36	1.Fh3 Rb6 2.Fg2 Ra6 3.Rb8 Rb6 4.Cg3 Rb5 5.Rc7 Rc5 6.Cf5 Rb4 7.Rd6 Rc4 8.Ff1+ Rc3 9.Rc5 Rb3 10.Ce3 Rb2 11.Rc4 Rc1 12.Rc3 Rb1 13.Cc2 Rc1 14.Fg2 Rb1 15.Fd5 Rc1 16.Fa2 Rd1 17.Cd4 Re1 18.Rd3 Rf2 19.Ce2 Rg2 20.Fe6 Rf3 21.Ff5 Rf2 22.Fg4 Re1 23.Cg3 Rf2 24.Cf5 Re1 25.Re3 Rf1 26.Ff3 Rg1 27.Ch4 Rf1 28.Fe2+ Rg1 29.Rf3 Rh1 30.Rf2 Rh2 31.Cf5 Rh3 32.Fh5 Rh2 33.Fg4 Rh1 34.Cg3+ Rh2 35.Cf1+ Rh1 36.Ff3+
 57	07:45	 6.907.116k	14.845k	+M36	1.Fh3 Rb6 2.Fg2 Ra6 3.Rb8 Rb6 4.Cg3 Rb5 5.Rc7 Rc5 6.Cf5 Rb4 7.Rd6 Rc4 8.Ff1+ Rc3 9.Rc5 Rb3 10.Ce3 Rb2 11.Rc4 Rc1 12.Rc3 Rb1 13.Cc2 Rc1 14.Fg2 Rb1 15.Fd5 Rc1 16.Fa2 Rd1 17.Cd4 Re1 18.Rd3 Rf2 19.Ce2 Rg2 20.Fe6 Rf3 21.Ff5 Rf2 22.Fg4 Re1 23.Cg3 Rf2 24.Cf5 Re1 25.Re3 Rf1 26.Ff3 Rg1 27.Ch4 Rf1 28.Fe2+ Rg1 29.Rf3 Rh1 30.Rf2 Rh2 31.Cf5 Rh3 32.Fh5 Rh2 33.Fg4 Rh1 34.Cg3+ Rh2 35.Cf1+ Rh1 36.Ff3+
 58	08:08	 7.236.702k	14.809k	+M36	1.Fh3 Rb6 2.Fg2 Ra6 3.Rb8 Rb6 4.Cg3 Rb5 5.Rc7 Rc5 6.Cf5 Rb4 7.Rd6 Rc4 8.Ff1+ Rc3 9.Rc5 Rb3 10.Ce3 Rb2 11.Rc4 Rc1 12.Rc3 Rb1 13.Cc2 Rc1 14.Fg2 Rb1 15.Fd5 Rc1 16.Fa2 Rd1 17.Cd4 Re1 18.Rd3 Rf2 19.Ce2 Rg2 20.Fe6 Rf3 21.Ff5 Rf2 22.Fg4 Re1 23.Cg3 Rf2 24.Cf5 Re1 25.Re3 Rf1 26.Ff3 Re1 27.Ch4 Rf1 28.Fe2+ Rg1 29.Rf3 Rh1 30.Rf2 Rh2 31.Cf5 Rh3 32.Fh5 Rh2 33.Fg4 Rh1 34.Cg3+ Rh2 35.Cf1+ Rh1 36.Ff3+
 59	08:39	 7.645.314k	14.720k	+M36	1.Fh3 Rb6 2.Fg2 Ra6 3.Rb8 Rb6 4.Cg3 Rb5 5.Rc7 Rc5 6.Cf5 Rb4 7.Rd6 Rc4 8.Ff1+ Rc3 9.Rc5 Rb3 10.Ce3 Rb2 11.Rc4 Rc1 12.Rc3 Rb1 13.Cc2 Rc1 14.Fg2 Rb1 15.Fd5 Rc1 16.Fa2 Rd1 17.Cd4 Re1 18.Rd3 Rf2 19.Ce2 Rg2 20.Fe6 Rf3 21.Ff5 Rf2 22.Fg4 Re1 23.Cg3 Rf2 24.Cf5 Re1 25.Re3 Rf1 26.Ff3 Re1 27.Ch4 Rf1 28.Fe2+ Rg1 29.Rf3 Rh1 30.Rf2 Rh2 31.Cf5 Rh3 32.Fh5 Rh2 33.Fg4 Rh1 34.Cg3+ Rh2 35.Cf1+ Rh1 36.Ff3+
 60	09:22	 8.243.703k	14.642k	+M36	1.Fh3 Rb6 2.Fg2 Ra6 3.Rb8 Rb6 4.Cg3 Rb5 5.Rc7 Rc5 6.Cf5 Rb4 7.Rd6 Rc4 8.Ff1+ Rc3 9.Rc5 Rb3 10.Ce3 Rb2 11.Rc4 Rc1 12.Rc3 Rb1 13.Cc2 Rc1 14.Fg2 Rb1 15.Fd5 Rc1 16.Fa2 Rd1 17.Cd4 Re1 18.Rd3 Rf2 19.Ce2 Rg2 20.Fe6 Rf3 21.Ff5 Rf2 22.Fg4 Re1 23.Cg3 Rf2 24.Cf5 Re1 25.Re3 Rf1 26.Ff3 Re1 27.Ch4 Rf1 28.Fe2+ Rg1 29.Rf3 Rh1 30.Rf2 Rh2 31.Cf5 Rh3 32.Fh5 Rh2 33.Fg4 Rh1 34.Cg3+ Rh2 35.Cf1+ Rh1 36.Ff3+
 61	09:48	 8.587.702k	14.583k	+M34	1.Fh3 Rb6 2.Fg2 Ra6 3.Rb8 Rb6 4.Cg3 Rb5 5.Rc7 Rc5 6.Cf5 Rb4 7.Rd6 Rc4 8.Ff1+ Rc3 9.Rc5 Rb3 10.Ce3 Rb2 11.Rc4 Rc1 12.Rc3 Rb1 13.Cc2 Rc1 14.Fg2 Rb1 15.Fd5 Rc1 16.Fa2 Rd1 17.Cd4 Re1 18.Fd5 Rd1 19.Fe4 Rc1 20.Cb3+ Rd1 21.Ff3+ Re1 22.Rd3 Rf2 23.Cd4 Rg3 24.Re3 Rh4 25.Rf4 Rh3 26.Cf5 Rh2 27.Re3 Rg1 28.Cg3 Rh2 29.Rf2 Rh3 30.Cf5 Rh2 31.Fg4 Rh1 32.Cg3+ Rh2 33.Cf1+ Rh1 34.Ff3+
 62	10:16	 8.968.608k	14.540k	+M34	1.Fh3 Rb6 2.Fg2 Ra6 3.Rb8 Rb6 4.Cg3 Rb5 5.Rc7 Rc5 6.Cf5 Rb4 7.Rd6 Rc4 8.Ff1+ Rc3 9.Rc5 Rb3 10.Ce3 Rb2 11.Rc4 Rc1 12.Rc3 Rb1 13.Cc2 Rc1 14.Fg2 Rb1 15.Fd5 Rc1 16.Fa2 Rd1 17.Cd4 Re1 18.Fd5 Rd1 19.Fe4 Rc1 20.Cb3+ Rd1 21.Ff3+ Re1 22.Rd3 Rf2 23.Cd4 Rg3 24.Re3 Rh4 25.Rf4 Rh3 26.Cf5 Rh2 27.Re3 Rg1 28.Cg3 Rh2 29.Rf2 Rh3 30.Cf5 Rh2 31.Fg4 Rh1 32.Cg3+ Rh2 33.Cf1+ Rh1 34.Ff3+
 63	11:06	 9.637.497k	14.470k	+M34	1.Fh3 Rb6 2.Fg2 Ra6 3.Rb8 Rb6 4.Cg3 Rb5 5.Rc7 Rc5 6.Cf5 Rb4 7.Rd6 Rc4 8.Ff1+ Rc3 9.Rc5 Rb3 10.Ce3 Rb2 11.Rc4 Rc1 12.Rc3 Rb1 13.Cc2 Rc1 14.Fg2 Rb1 15.Fd5 Rc1 16.Fa2 Rd1 17.Cd4 Re1 18.Fd5 Rd1 19.Fe4 Rc1 20.Cb3+ Rd1 21.Ff3+ Re1 22.Rd3 Rf2 23.Cd4 Rg3 24.Re3 Rh4 25.Rf4 Rh3 26.Cf5 Rh2 27.Re3 Rg1 28.Cg3 Rh2 29.Rf2 Rh3 30.Cf5 Rh2 31.Fg4 Rh1 32.Cg3+ Rh2 33.Cf1+ Rh1 34.Ff3+
 64	12:12	 10.494.911k	14.332k	+M34	1.Fh3 Rb6 2.Fg2 Ra6 3.Rb8 Rb6 4.Cg3 Rb5 5.Rc7 Rc5 6.Cf5 Rb4 7.Rd6 Rc4 8.Ff1+ Rc3 9.Rc5 Rb3 10.Ce3 Rb2 11.Rc4 Rc1 12.Rc3 Rb1 13.Cc2 Rc1 14.Fg2 Rb1 15.Fd5 Rc1 16.Fa2 Rd1 17.Cd4 Re1 18.Fd5 Rd1 19.Fe4 Rc1 20.Cb3+ Rd1 21.Ff3+ Re1 22.Rd3 Rf2 23.Cd4 Rg3 24.Re3 Rh4 25.Rf4 Rh3 26.Cf5 Rh2 27.Re3 Rg1 28.Cg3 Rh2 29.Rf2 Rh3 30.Cf5 Rh2 31.Fg4 Rh1 32.Cg3+ Rh2 33.Cf1+ Rh1 34.Ff3+
 65	13:35	 11.526.235k	14.130k	+M34	1.Fh3 Rb6 2.Fg2 Ra6 3.Rb8 Rb6 4.Cg3 Rb5 5.Rc7 Rc5 6.Cf5 Rb4 7.Rd6 Rc4 8.Ff1+ Rc3 9.Rc5 Rb3 10.Ce3 Rb2 11.Rc4 Rc1 12.Rc3 Rb1 13.Cc2 Rc1 14.Fg2 Rb1 15.Fd5 Rc1 16.Fa2 Rd1 17.Cd4 Re1 18.Fd5 Rd1 19.Fe4 Rc1 20.Cb3+ Rd1 21.Ff3+ Re1 22.Rd3 Rf2 23.Cd4 Rf1 24.Re3 Re1 25.Cc6 Rf1 26.Cb4 Rg1 27.Cd5 Rf1 28.Fe2+ Rg2 29.Ce7 Rg3 30.Cg6 Rg2 31.Cf4+ Rg3 32.Ce6 Rh3 33.Ff3 Rg3 34.Cd4 Rh4 35.Rf4 Rh3 36.Cc2 Rh4 37.Ce3 Rh3 38.Fg4+ Rh2 39.Rf3 Rg1 40.Rg3 Rh1 41.Rf2 Rh2 42.Cf1+ Rh1 43.Ff3+
 66	14:39	 12.365.300k	14.065k	+M33	1.Fh3 Rb6 2.Fg2 Ra6 3.Rb8 Rb6 4.Cg3 Rb5 5.Rc7 Rc5 6.Cf5 Rb4 7.Rd6 Rc4 8.Ff1+ Rc3 9.Rc5 Rb3 10.Ce3 Rb2 11.Rc4 Rc1 12.Rc3 Rb1 13.Cc2 Rc1 14.Fg2 Rb1 15.Fd5 Rc1 16.Fa2 Rd1 17.Cd4 Re1 18.Fd5 Rd1 19.Fe4 Rc1 20.Cb3+ Rd1 21.Ff3+ Re1 22.Rd3 Rf2 23.Cd4 Rf1 24.Re3 Re1 25.Cc2+ Rf1 26.Fe4 Rg1 27.Rf3 Rh2 28.Ff5 Rg1 29.Ce3 Rh2 30.Rf2 Rh1 31.Fd7 Rh2 32.Cf1+ Rh1 33.Fc6+
 67	15:51	 13.416.591k	14.093k	+M33	1.Fh3 Rb6 2.Fg2 Ra6 3.Rb8 Rb6 4.Cg3 Rb5 5.Rc7 Rc5 6.Cf5 Rb4 7.Rd6 Rc4 8.Ff1+ Rc3 9.Rc5 Rb3 10.Ce3 Rb2 11.Rc4 Rc1 12.Rc3 Rb1 13.Cc2 Rc1 14.Fg2 Rb1 15.Fd5 Rc1 16.Fa2 Rd1 17.Cd4 Re1 18.Fd5 Rd1 19.Fe4 Rc1 20.Cb3+ Rd1 21.Ff3+ Re1 22.Rd3 Rf2 23.Cd4 Rf1 24.Re3 Re1 25.Cc2+ Rf1 26.Fe4 Rg1 27.Rf3 Rh2 28.Ff5 Rg1 29.Ce3 Rh2 30.Rf2 Rh1 31.Fd7 Rh2 32.Cf1+ Rh1 33.Fc6+
 68	16:52	 14.295.580k	14.115k	+M33	1.Fh3 Rb6 2.Fg2 Ra6 3.Rb8 Rb6 4.Cg3 Rb5 5.Rc7 Rc5 6.Cf5 Rb4 7.Rd6 Rc4 8.Ff1+ Rc3 9.Rc5 Rb3 10.Ce3 Rb2 11.Rc4 Rc1 12.Rc3 Rb1 13.Cc2 Rc1 14.Fg2 Rb1 15.Fd5 Rc1 16.Fa2 Rd1 17.Cd4 Re1 18.Fd5 Rd1 19.Fe4 Rc1 20.Cb3+ Rd1 21.Ff3+ Re1 22.Rd3 Rf2 23.Cd4 Rf1 24.Re3 Re1 25.Cc2+ Rf1 26.Fe4 Rg1 27.Rf3 Rh2 28.Ff5 Rg1 29.Ce3 Rh2 30.Rf2 Rh1 31.Fd7 Rh2 32.Cf1+ Rh1 33.Fc6+
 69	18:53	 16.044.015k	14.157k	+M33	1.Fh3 Rb6 2.Fg2 Ra6 3.Rb8 Rb6 4.Cg3 Rb5 5.Rc7 Rc5 6.Cf5 Rb4 7.Rd6 Rc4 8.Ff1+ Rc3 9.Rc5 Rb3 10.Ce3 Rb2 11.Rc4 Rc1 12.Rc3 Rb1 13.Cc2 Rc1 14.Fg2 Rb1 15.Fd5 Rc1 16.Fa2 Rd1 17.Cd4 Re1 18.Fd5 Rd1 19.Fe4 Rc1 20.Cb3+ Rd1 21.Ff3+ Re1 22.Rd3 Rf2 23.Cd4 Rf1 24.Re3 Re1 25.Cc2+ Rf1 26.Fe4 Rg1 27.Rf3 Rh2 28.Ff5 Rg1 29.Ce3 Rh2 30.Rf2 Rh1 31.Fd7 Rh2 32.Cf1+ Rh1 33.Fc6+
 70	20:31	 17.492.838k	14.207k	+M33	1.Fh3 Rb6 2.Fg2 Ra6 3.Rb8 Rb6 4.Cg3 Rb5 5.Rc7 Rc5 6.Cf5 Rb4 7.Rd6 Rc4 8.Ff1+ Rc3 9.Rc5 Rb3 10.Ce3 Rb2 11.Rc4 Rc1 12.Rc3 Rb1 13.Cc2 Rc1 14.Fg2 Rb1 15.Fd5 Rc1 16.Fa2 Rd1 17.Cd4 Re1 18.Fd5 Rd1 19.Fe4 Rc1 20.Cb3+ Rd1 21.Ff3+ Re1 22.Rd3 Rf2 23.Cd4 Rf1 24.Re3 Re1 25.Cc2+ Rf1 26.Fe4 Rg1 27.Rf3 Rh2 28.Ff5 Rg1 29.Ce3 Rh2 30.Rf2 Rh1 31.Fd7 Rh2 32.Cf1+ Rh1 33.Fc6+

duncan
Posts: 12031
Joined: Mon Jul 07, 2008 8:50 pm

Re: Stockfish version with hash saving capability

Post by duncan » Mon Aug 21, 2017 4:04 pm

zullil wrote:
duncan wrote:
zullil wrote:
duncan wrote:
zullil wrote: asmFish eventually found mate-in-11 from the original position. But it took a while.

Code: Select all

info depth 54 seldepth 24 multipv 1 time 5491130 nps 50956798 score mate 11 nodes 279810404462 hashfull 999 tbhits 0 pv a2a4 h7h5 d2d4 c7c6 e2e4 g7g6 d4d5 g6g5 d1h5 e8d8 h5f7 c6d5 f1b5 d5e4 f7d5 d8c8 d5d7 c8b8 b5a6 b8a8 d7b7

is your nps 50956798 which is very fast and quicker than threadripper.? May I ask what sytem you are using?
Not sure what your first sentence is trying to ask. :wink: Yes, the nodes/sec is 50956798, which is indeed very fast. I don't know anything about Threadripper. I have 2 x Intel(R) Xeon(R) CPU E5-2687W v3 @ 3.10GHz and 64GB RAM. I tend to use asmFish because it is NUMA-aware, which seems to improve performance on my system.
the following position is a mater in 33 and is the Longest mate in King, Bishop and Knight versus King endgame.

it takes me a long time on my computer to find the mate in 33. how long do you estimate it should take on your computer .just 10 minutes ?

[d]K7/2kB4/8/8/8/8/8/5N2 w - - 0 1

http://www.gilith.com/chess/endgames/kbn_k.html
The time until mate-in-33 is reported varies quite a bit, due to the randomness inherent in multithreaded searches. The shortest time was about 1.5 minutes and the longest was about 5 minutes. The following seems typical:

Code: Select all

info depth 64 seldepth 72 multipv 1 time 129908 nps 73529464 score mate 33 nodes 9552065660 hashfull 67 tbhits 0 pv d7h3 c7b6 h3g2 b6c5 a8b7 c5d6 b7b6 d6e6 b6c6 e6e5 g2d5 e5f4 c6d6 f4f5 f1e3 f5f4 e3c4 f4g5 d6e7 g5g6 d5e4 g6g7 c4e5 g7g8 e7f6 g8h8 e4d5 h8h7 e5g6 h7h6 d5g8 h6h5 g6e5 h5h4 g8d5 h4g3 e5d3 g3g4 f6g6 g4h4 d5f3 h4g3 f3d1 g3g2 g6g5 g2g3 d1e2 g3h3 g5f4 h3h4 d3e1 h4h3 e2g4 h3h2 f4f3 h2h1 g4h3 h1h2 f3g4 h2g1 g4g3 g1h1 h3g2 h1g1 e1f3
This was done using a 4 GB hash table. The size of the hash will also affect the solution time.
thanks.

1.5 minutes. super fast. was not expecting that.
you found mate in 33 with 64 ply search, Vincent took longer with 66 ply. is this because of the randomness inherent in multithreaded searches. ?

also time 129908 means what ?

duncan
Posts: 12031
Joined: Mon Jul 07, 2008 8:50 pm

Re: Stockfish version with hash saving capability

Post by duncan » Mon Aug 21, 2017 4:32 pm

Vinvin wrote:
Sting-SF 8.5, 4*4GHz. Mate in 60 in 4m17s, mate in 33 in 14m39s.

thanks for running it. does sting-sf have an advantage over normal sf ?

also how much hash did you use ?

Vinvin
Posts: 4919
Joined: Thu Mar 09, 2006 8:40 am
Full name: Vincent Lejeune

Re: Stockfish version with hash saving capability

Post by Vinvin » Mon Aug 21, 2017 4:57 pm

duncan wrote: also time 129908 means what ?
This kind of output is in milliseconds unit, so : about 130 seconds.

Vinvin
Posts: 4919
Joined: Thu Mar 09, 2006 8:40 am
Full name: Vincent Lejeune

Re: Stockfish version with hash saving capability

Post by Vinvin » Mon Aug 21, 2017 5:25 pm

duncan wrote:
Vinvin wrote:
Sting-SF 8.5, 4*4GHz. Mate in 60 in 4m17s, mate in 33 in 14m39s.

thanks for running it. does sting-sf have an advantage over normal sf ?

also how much hash did you use ?
I don't know what are the mods compare to the regular SF but Sting-SF can find very long mates.

Post Reply