Page 1 of 2
Andscacs test version
Posted: Sun Dec 09, 2018 12:47 pm
by cdani
Hello!
Maybe someone can test if this test version of Andscacs scales well in nps with more than 64 threads in Windows?
Also included is a Linux version that should work well, using events instead of sleep for the waiting threads (Windows version already had this for quite some time), and maybe someone can test it also, better with a lot of threads
.
http://www.andscacs.com/downloads/andscacs094157.zip
This test version should be more than 20 elo stronger in selfplay than 0.94.
Thanks!!
Re: Andscacs test version
Posted: Sun Dec 09, 2018 8:51 pm
by cdani
The linux version was Ubuntu. I added to the same file a Centos 7 version. Otherwise it said GLIBC... not found.
Re: Andscacs test version
Posted: Mon Dec 10, 2018 4:35 am
by arunsoorya1309
we are testing in ccc to move to Linux, I’ll try to get it in
Regards
Arun
Re: Andscacs test version
Posted: Mon Dec 10, 2018 7:08 am
by cdani
arunsoorya1309 wrote: ↑Mon Dec 10, 2018 4:35 am
we are testing in ccc to move to Linux, I’ll try to get it in
Regards
Arun
Thanks! Please, tell me something a little before starting the tournament, as I will send you my latest version.
Re: Andscacs test version
Posted: Mon Dec 10, 2018 11:31 am
by Werner
Hi Daniel,
I can only show you the results between 1CPU and 8Threads (4CPU):
Ver 0.94 1CPU 512 MB hash T30
Startposition 142 sec +0.22 e4 1318nps
after Nh3 with black 204 sec -0.43 d5 1302 nps
Ver 0.94 8Thr 2048 MB hash T30
Startposition 114 sec +0.32 e4 6812 nps
after Nh3 with black 120 sec -0.44 d5 6592 nps
Ver 0.94157 1CPU 512 MB hash T30
Startposition 345 sec +0.23 e4 1314nps
after Nh3 with black 208 sec -0.43 d5 1313 nps
Ver 0.94157 8Thr 2048 MB hash T30
Startposition 224 sec +0.24 e4 6819 nps
after Nh3 with black 172 sec -0.42 d5 6621 nps
Can we soon have a new version for testing?
Re: Andscacs test version
Posted: Mon Dec 10, 2018 8:36 pm
by cdani
Werner wrote: ↑Mon Dec 10, 2018 11:31 am
Hi Daniel,
I can only show you the results between 1CPU and 8Threads (4CPU):
Ver 0.94 1CPU 512 MB hash T30
Startposition 142 sec +0.22 e4 1318nps
after Nh3 with black 204 sec -0.43 d5 1302 nps
Ver 0.94 8Thr 2048 MB hash T30
Startposition 114 sec +0.32 e4 6812 nps
after Nh3 with black 120 sec -0.44 d5 6592 nps
Ver 0.94157 1CPU 512 MB hash T30
Startposition 345 sec +0.23 e4 1314nps
after Nh3 with black 208 sec -0.43 d5 1313 nps
Ver 0.94157 8Thr 2048 MB hash T30
Startposition 224 sec +0.24 e4 6819 nps
after Nh3 with black 172 sec -0.42 d5 6621 nps
Can we soon have a new version for testing?
Thanks!
In a few weeks I will publish a new "official" version.
Re: Andscacs test version
Posted: Tue Dec 11, 2018 7:24 am
by noobpwnftw
I have tested the CentOS version and seems to scale without problem with 128 cores.
Re: Andscacs test version
Posted: Tue Dec 11, 2018 8:29 am
by Gary Internet
cdani wrote: ↑Mon Dec 10, 2018 8:36 pm
Thanks!
In a few weeks I will publish a new "official" version.
Does this mean we will see Andscacs 0.95 playing in the TCEC Premier Division?
Re: Andscacs test version
Posted: Tue Dec 11, 2018 1:08 pm
by Eduard
Will it be possible for the engine to use endgame tablebases?
Re: Andscacs test version
Posted: Tue Dec 11, 2018 2:48 pm
by cdani
noobpwnftw wrote: ↑Tue Dec 11, 2018 7:24 am
I have tested the CentOS version and seems to scale without problem with 128 cores.
Many thanks!!
I also see that it's in TCEC, analyzing as a second engine, nice
Gary Internet wrote: ↑Tue Dec 11, 2018 8:29 am
cdani wrote: ↑Mon Dec 10, 2018 8:36 pm
Thanks!
In a few weeks I will publish a new "official" version.
Does this mean we will see Andscacs 0.95 playing in the TCEC Premier Division?
Yes, is the most probable.
Eduard wrote: ↑Tue Dec 11, 2018 1:08 pm
Will it be possible for the engine to use endgame tablebases?
Will be nice, but for the moment I don't want to spend time on this, as I have too much other ideas to try.