Before I head to get some sleep, I will begin a 100 game match at 4'+2" for a time control- between Dmitri's newest Firenzina version and this latest Gull release, version R375 x64. As for Gull, I have given it a good bit of exposure, but have not yet run a match with one of Dmitri's versions. He has worked hard, and I think he deserves to have his engine publicly tested. The controls are as follows:
XPS 8500 Intel i5-4 True Cores
Fritz 11 gui
1CPU/64bit
128MB hash
Bases=NONE
Ponder_Learning=OFF
Perfect 2012b.ctg w/12-move limit
4'+2"
Match=100 games
Firenzina 2.2.2 xTreme x64 pcnt v Gull R375 x64
I would not be one to question a programmer, but I just cannot understand why ThinkingALot seems to be bound and determined to keep his hash size so low. Oh well.
Updates will be given-
george
An Interesting New Match That I Will Begin Very Shortly!
Moderator: Ras
-
geots
- Posts: 4790
- Joined: Sat Mar 11, 2006 12:42 am
-
velmarin
- Posts: 1600
- Joined: Mon Feb 21, 2011 9:48 am
Re: An Interesting New Match That I Will Begin Very Shortly!
Gull R375
fprintf(stdout,"option name Hash type spin min 1 max 65536 default 16\n");
Minimo 1 Megabyte
maximo 65536 Megabyte
default 16 Megabyte
fprintf(stdout,"option name Hash type spin min 1 max 65536 default 16\n");
Minimo 1 Megabyte
maximo 65536 Megabyte
default 16 Megabyte
-
geots
- Posts: 4790
- Joined: Sat Mar 11, 2006 12:42 am
Re: An Interesting New Match That I Will Begin Very Shortly!
velmarin wrote:Gull R375
fprintf(stdout,"option name Hash type spin min 1 max 65536 default 16\n");
Minimo 1 Megabyte
maximo 65536 Megabyte![]()
default 16 Megabyte
Velmarin, I see some figures I recognize- but I just don't have any idea the point you are trying to make. At first I thought it had to do with changing the hash on my end- but running Gull in Fritz 11 that is not possible. At least as far as I know.
Best,
george
-
Ajedrecista
- Posts: 2176
- Joined: Wed Jul 13, 2011 9:04 pm
- Location: Madrid, Spain.
Re: An Interesting New Match That I Will Begin Very Shortly!
Hi George:
1 TB of maximum hash! It is incredible for me.
Good luck with the match. My best to you.
Regards from Spain.
Ajedrecista.
I think he is impressed with the big size of the maximum hash: 65536 MB (64 GB). But Zappa Mexico II is the king:geots wrote:velmarin wrote:Gull R375
fprintf(stdout,"option name Hash type spin min 1 max 65536 default 16\n");
Minimo 1 Megabyte
maximo 65536 Megabyte![]()
default 16 Megabyte
Velmarin, I see some figures I recognize- but I just don't have any idea the point you are trying to make. At first I thought it had to do with changing the hash on my end- but running Gull in Fritz 11 that is not possible. At least as far as I know.
Best,
george
Code: Select all
uci
id name Zappa Mexico II
[...]
option name Hash type spin default 128 min 4 max 1048576
[...]
uciokGood luck with the match. My best to you.
Regards from Spain.
Ajedrecista.
-
geots
- Posts: 4790
- Joined: Sat Mar 11, 2006 12:42 am
Re: An Interesting New Match That I Will Begin Very Shortly!
Ajedrecista wrote:Hi George:
I think he is impressed with the big size of the maximum hash: 65536 MB (64 GB). But Zappa Mexico II is the king:geots wrote:velmarin wrote:Gull R375
fprintf(stdout,"option name Hash type spin min 1 max 65536 default 16\n");
Minimo 1 Megabyte
maximo 65536 Megabyte![]()
default 16 Megabyte
Velmarin, I see some figures I recognize- but I just don't have any idea the point you are trying to make. At first I thought it had to do with changing the hash on my end- but running Gull in Fritz 11 that is not possible. At least as far as I know.
Best,
george
1 TB of maximum hash! It is incredible for me.Code: Select all
uci id name Zappa Mexico II [...] option name Hash type spin default 128 min 4 max 1048576 [...] uciok
Good luck with the match. My best to you.
Regards from Spain.
Ajedrecista.
Jesus, I would imagine he could say the maximum hash is 100 TB. Or anything else he wanted to say. Because it is running around 30,000 k memory, which means less than 30MB hash- and you cannot change it to anything higher. When you can't disprove a statement- it's ok to say anything.
Regards,
george
-
Gusev
- Posts: 1476
- Joined: Mon Jan 28, 2013 2:51 pm
Re: An Interesting New Match That I Will Begin Very Shortly!
Just a wild guess. Given that Vadim's parallelism is implemented via processes, not threads, could this be 30MB per process, yet reported this way? With 8 processes, this makes 240MB for all of them taken together.Because it is running around 30,000 k memory, which means less than 30MB hash- and you cannot change it to anything higher.
Is R375 publicly available yet?
-
Ajedrecista
- Posts: 2176
- Joined: Wed Jul 13, 2011 9:04 pm
- Location: Madrid, Spain.
Re: An Interesting New Match That I Will Begin Very Shortly!
Hello Dmitri:
http://www.sdchess.ru/download_engines.htm
Gull R375+sources
The improvement is almost not measurable:
Gull
Ajedrecista.
Yes it is. Take a look here:Gusev wrote:Is R375 publicly available yet?
http://www.sdchess.ru/download_engines.htm
Gull R375+sources
The improvement is almost not measurable:
Gull
Regards from Spain.Kevin wrote:Gull R375 is available, +5 elo
Ajedrecista.
-
Gusev
- Posts: 1476
- Joined: Mon Jan 28, 2013 2:51 pm
Re: An Interesting New Match That I Will Begin Very Shortly!
Thank you very much, I have downloaded R375. At TCEC, even +5 Elo might matter. It's like one of those legendary races of steamers on Mississippi River. 
-
velmarin
- Posts: 1600
- Joined: Mon Feb 21, 2011 9:48 am
Re: An Interesting New Match That I Will Begin Very Shortly!
Dmitri
In this Wiki are always news, download links, ect.
Always very interesting, you have your Firenzina.
http://computer-chess.org/doku.php?id=c ... ngine_list
Pablo,
Geots dijo
Saludos.
In this Wiki are always news, download links, ect.
Always very interesting, you have your Firenzina.
http://computer-chess.org/doku.php?id=c ... ngine_list
Pablo,
Geots dijo
Code: Select all
no estaría a cuestionar un programador, pero no puedo entender por qué ThinkingALot parece estar atado y decidido a mantener su tamaño de hash tan bajo.Code: Select all
Lo mio sólo fue una aclaración de hasta donde llegan las posibilidades de GULL, en Hash, 1024 para 32bits, 65.536 para 64 bits.
Otra cosa es no saber poner el numero en fritz.-
Gusev
- Posts: 1476
- Joined: Mon Jan 28, 2013 2:51 pm
Re: An Interesting New Match That I Will Begin Very Shortly!
I have bookmarked this page, thank you! I see Fire on the list of "dead engines" there. Indeed, who would want to test a dead engine? But, as we see, with open source, a good engine can be "undead".In this Wiki are always news, download links, ect.
Always very interesting, you have your Firenzina.
http://computer-chess.org/doku.php?id=c ... ngine_list