Leela Chess Zero 42565 vs Stockfish 140619
Moderators: hgm, Rebel, chrisw
-
- Posts: 2821
- Joined: Fri Sep 25, 2015 9:38 pm
- Location: Sortland, Norway
Re: Leela Chess Zero 42565 vs Stockfish 140619
mwyoung: why can't you just simply adjust threads as per general consensus and everything is fine.
-
- Posts: 2727
- Joined: Wed May 12, 2010 10:00 pm
Re: Leela Chess Zero 42565 vs Stockfish 140619
You asked a half way respectabel question. Other then you think I need some kind community approval for my settings.Nordlandia wrote: ↑Sun Jun 16, 2019 7:04 pm mwyoung: why can't you just simply adjust threads as per general consensus and everything is fine.
Because they are already adjusted. I did it in the most efficient way. Not the hack slow caveman way the community suggest.
Last edited by mwyoung on Sun Jun 16, 2019 7:38 pm, edited 1 time in total.
"The worst thing that can happen to a forum is a running wild attacking moderator(HGM) who is not corrected by the community." - Ed Schröder
But my words like silent raindrops fell. And echoed in the wells of silence.
But my words like silent raindrops fell. And echoed in the wells of silence.
-
- Posts: 10948
- Joined: Wed Jul 26, 2006 10:21 pm
- Full name: Kai Laskos
Re: Leela Chess Zero 42565 vs Stockfish 140619
Leave this useless test and thread, here is one proceeding right now using similar hardware, but in correct conditions (ponder = on):Nordlandia wrote: ↑Sun Jun 16, 2019 7:04 pm mwyoung: why can't you just simply adjust threads as per general consensus and everything is fine.
viewtopic.php?f=6&t=70519&start=20Hugo wrote: ↑Sun Jun 16, 2019 1:06 pm new testrun started with nn 42578
16 core XEON v4 PC, HT off
12 cpu SF
2cpu Lc0
5m+3s ponder ON
180 games, will take 3 days
https://docs.google.com/spreadsheets/d/ ... sp=sharing
regards
C.K.
-
- Posts: 2727
- Joined: Wed May 12, 2010 10:00 pm
Re: Leela Chess Zero 42565 vs Stockfish 140619
The last time I checked no one forced you here to this thread. So don't let the door hit the troll in the ass.Laskos wrote: ↑Sun Jun 16, 2019 7:19 pmLeave this useless test and thread, here is one proceeding right now using similar hardware, but in correct conditions (ponder = on):Nordlandia wrote: ↑Sun Jun 16, 2019 7:04 pm mwyoung: why can't you just simply adjust threads as per general consensus and everything is fine.
viewtopic.php?f=6&t=70519&start=20Hugo wrote: ↑Sun Jun 16, 2019 1:06 pm new testrun started with nn 42578
16 core XEON v4 PC, HT off
12 cpu SF
2cpu Lc0
5m+3s ponder ON
180 games, will take 3 days
https://docs.google.com/spreadsheets/d/ ... sp=sharing
regards
C.K.
"The worst thing that can happen to a forum is a running wild attacking moderator(HGM) who is not corrected by the community." - Ed Schröder
But my words like silent raindrops fell. And echoed in the wells of silence.
But my words like silent raindrops fell. And echoed in the wells of silence.
-
- Posts: 2821
- Joined: Fri Sep 25, 2015 9:38 pm
- Location: Sortland, Norway
Re: Leela Chess Zero 42565 vs Stockfish 140619
Laskos: i guess it's not possible to follow that match or is it?
-
- Posts: 10948
- Joined: Wed Jul 26, 2006 10:21 pm
- Full name: Kai Laskos
Re: Leela Chess Zero 42565 vs Stockfish 140619
No, I don't think it's possible, only the results as they come.Nordlandia wrote: ↑Sun Jun 16, 2019 7:41 pm Laskos: i guess it's not possible to follow that match or is it?
-
- Posts: 2821
- Joined: Fri Sep 25, 2015 9:38 pm
- Location: Sortland, Norway
Re: Leela Chess Zero 42565 vs Stockfish 140619
Wery well. I PM'ed Graham whether he'll try ponder on his dual xeon motherboard. Graham's engine games is watchable in playchess server.
Permanent brain enabled make a lot more sense on an dual socket mobo than single cpu motherboard.
Permanent brain enabled make a lot more sense on an dual socket mobo than single cpu motherboard.
-
- Posts: 2727
- Joined: Wed May 12, 2010 10:00 pm
Re: Leela Chess Zero 42565 vs Stockfish 140619
That is not correct. But dual socket is still ok. AMD 2950x and above is a dual or quad chip design connected with infinity fabric. This innovation was designed to speed up the inherent latency issues in the dual socket mother board design.Nordlandia wrote: ↑Sun Jun 16, 2019 7:54 pm Wery well. I PM'ed Graham whether he'll try ponder on his dual xeon motherboard. Graham's engine games is watchable in playchess server.
Permanent brain enabled make a lot more sense on an dual socket mobo than single cpu motherboard.
"The worst thing that can happen to a forum is a running wild attacking moderator(HGM) who is not corrected by the community." - Ed Schröder
But my words like silent raindrops fell. And echoed in the wells of silence.
But my words like silent raindrops fell. And echoed in the wells of silence.
-
- Posts: 2821
- Joined: Fri Sep 25, 2015 9:38 pm
- Location: Sortland, Norway
Re: Leela Chess Zero 42565 vs Stockfish 140619
Then again the question that arise is if ponder is good trade-off for more time
-
- Posts: 27817
- Joined: Fri Mar 10, 2006 10:06 am
- Location: Amsterdam
- Full name: H G Muller
Re: Leela Chess Zero 42565 vs Stockfish 140619
Pondering makes sense when playing against Leela on a many-core machine, as Leela wouldn't use many CPU threads while thinking, and all other threads would then go to waste. Likewise the GPUs would be idle during Stockfish' turn when Leela is not pondering.
If Leela need two unshared cores, you can just set an affinity for it for hyper-threads 0-3. You can then set affinity for Stockfish to the remaining HT. That way they won't compete for cores. They might still compete for memory bandwidth, though; not sure how important that is for Leela.
I don't see why Leela couldn't run its two threads on the same physical core. That could of course be a disadvantage compared to running 2 threads on 2 physical cores, but that also holds for Stockfish' threads sharing physical cores. Yet some people claim that hyper-threading is beneficial compared to running 1 thread per physical core, and testing both under conditions with 2 active HT per core does not seem particularly unfair. It is just like all HT are somewhat slower physical cores. This doesn't involve any scheduling, so it should not be noisy.
Of course if you don't also reserve some cores for the OS, that would cause noise.
If Leela need two unshared cores, you can just set an affinity for it for hyper-threads 0-3. You can then set affinity for Stockfish to the remaining HT. That way they won't compete for cores. They might still compete for memory bandwidth, though; not sure how important that is for Leela.
I don't see why Leela couldn't run its two threads on the same physical core. That could of course be a disadvantage compared to running 2 threads on 2 physical cores, but that also holds for Stockfish' threads sharing physical cores. Yet some people claim that hyper-threading is beneficial compared to running 1 thread per physical core, and testing both under conditions with 2 active HT per core does not seem particularly unfair. It is just like all HT are somewhat slower physical cores. This doesn't involve any scheduling, so it should not be noisy.
Of course if you don't also reserve some cores for the OS, that would cause noise.