Another Stockfish Experiment

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

Moderator: Ras

Werewolf
Posts: 2054
Joined: Thu Sep 18, 2008 10:24 pm

Another Stockfish Experiment

Post by Werewolf »

I plan to run a this match:

SF 15 on 8 cores V SF 15 1 core, with the 1 core version getting 8x more time. Ponder off.

Other than predicting the 1 core version will win, anyone want to predict the elo difference?

(Openings will probably be Noomen Short)
Uri Blass
Posts: 10918
Joined: Thu Mar 09, 2006 12:37 am
Location: Tel-Aviv Israel

Re: Another Stockfish Experiment

Post by Uri Blass »

The question is what time control.

I am not sure if 1 core win and maybe we are going to see 100% draws.
Werewolf
Posts: 2054
Joined: Thu Sep 18, 2008 10:24 pm

Re: Another Stockfish Experiment

Post by Werewolf »

Maybe 1 min + 1s and 8 min + 8s
kasinp
Posts: 261
Joined: Sat Dec 02, 2006 10:47 pm
Location: Toronto
Full name: Peter Kasinski

Re: Another Stockfish Experiment

Post by kasinp »

Werewolf wrote: Wed May 04, 2022 7:27 pm I plan to run a this match:

SF 15 on 8 cores V SF 15 1 core, with the 1 core version getting 8x more time. Ponder off.

Other than predicting the 1 core version will win, anyone want to predict the elo difference?

(Openings will probably be Noomen Short)
I would go with +30 Elo for the 1CPU version with 8 times the clock.
Uri Blass
Posts: 10918
Joined: Thu Mar 09, 2006 12:37 am
Location: Tel-Aviv Israel

Re: Another Stockfish Experiment

Post by Uri Blass »

kasinp wrote: Thu May 05, 2022 12:55 am
Werewolf wrote: Wed May 04, 2022 7:27 pm I plan to run a this match:

SF 15 on 8 cores V SF 15 1 core, with the 1 core version getting 8x more time. Ponder off.

Other than predicting the 1 core version will win, anyone want to predict the elo difference?

(Openings will probably be Noomen Short)
I would go with +30 Elo for the 1CPU version with 8 times the clock.
First interesting question is of
SF 15 on 1 cores(8+8 time control) V SF 15 1 core(1+1 time control), opening by Noomen Short

Assuming you do not use some biased opening book I guess we are going to see mainly draws so I will not be surprised if you are going to get less than 30 elo even in this case.
Jouni
Posts: 3703
Joined: Wed Mar 08, 2006 8:15 pm
Full name: Jouni Uski

Re: Another Stockfish Experiment

Post by Jouni »

Waiting for results. SF wiki has one test:

Playing 8 threads vs 1 thread at LTC (60+0.6, 8moves_v3.pgn):

Score of t8 vs seq: 476 - 3 - 521 [0.737] 1000
Elo difference: 178.6 +/- 14.0, LOS: 100.0 %, DrawRatio: 52.1 %

Playing 1 thread at 8xLTC (480+4.8) vs (60+0.6) (8moves_v3.pgn):

Score of seq8 vs seq: 561 - 5 - 434 [0.778] 1000
Elo difference: 217.9 +/- 15.8, LOS: 100.0 %, DrawRatio: 43.4 %

Note this is pre-NNUE :D .
Jouni
Werewolf
Posts: 2054
Joined: Thu Sep 18, 2008 10:24 pm

Re: Another Stockfish Experiment

Post by Werewolf »

Sorry I've got covid :(

When I'm better I'll look into this
User avatar
Cumnor
Posts: 116
Joined: Fri Jan 27, 2012 4:23 pm
Location: Cumnor, Oxford, UK
Full name: Kevin D Plant

Re: Another Stockfish Experiment

Post by Cumnor »

Werewolf wrote: Wed May 18, 2022 6:41 pm Sorry I've got covid :(

When I'm better I'll look into this
Sorry to hear that, I wish you a speedy recovery.
Moderator of Rybka forum (Site no longer active)
Admin of Infinitychess playing server and Forum (Site suspended, maybe be back in the Future)