Leo wrote: ↑Sat Nov 20, 2021 10:59 pm
I found this information a few years ago. Quote:
EGTB
6 man tablebase syzygy {EGTB} 90 GB
7 man tablebase The high speed of generating the tablebases was because of using a supercomputer named Lomonosov (top500). The size of seven-man tablebases is about 140 TB.
Many show interest in what is to expect from 8-man endings. First, take note that the longest 6-man mate took 262 moves (KRN-KNN). Moving to 7-man endings doubled this value. Second, 8-man tablebases include much more endings with both sides having relatively equal strength. All this gives us a strong hope to discover a mate in more than 1000 moves in one of 8-man endgames. Unfortunately the size of 8-man tablebases will be 100 times larger than the size of 7-man tablebases. To fully compute them, one will need about 10 PB (10,000 TB) of disk space and 50 TB of RAM. Only the top 10 supercomputers can solve the 8-man problem in 2014. The first 1000-move mate is unlikely to be found until 2020 when a part of a TOP100 supercomputer may be allowed to be used for solving this task.
This information is outdated. https://syzygy-tables.info/ the 7 man Tablebase uses 18 Tbit which is around 0.33bit PER POSITION!
So you today can download and lookup a 7 man TB for around 2k dollars with current nvme ssd prices.
Yes the size of 8 man tablebase will be 110x larger in every domain - the point is that 8 man can be solved in 2022 as hardware has come a long way. Especially on the SSD side you can directly map ssd to gpu memory with directStorage and Cuda can access that.
Im not interested in x-to mate problem. What interests me is the ELO scaling from these tablebases as of today it seems that 6man tb is as good as one weak pruning commit for SF.
For 7 man I lack data.
Leo wrote: ↑Sat Nov 20, 2021 10:59 pm
I found this information a few years ago. Quote:
EGTB
6 man tablebase syzygy {EGTB} 90 GB
7 man tablebase The high speed of generating the tablebases was because of using a supercomputer named Lomonosov (top500). The size of seven-man tablebases is about 140 TB.
Many show interest in what is to expect from 8-man endings. First, take note that the longest 6-man mate took 262 moves (KRN-KNN). Moving to 7-man endings doubled this value. Second, 8-man tablebases include much more endings with both sides having relatively equal strength. All this gives us a strong hope to discover a mate in more than 1000 moves in one of 8-man endgames. Unfortunately the size of 8-man tablebases will be 100 times larger than the size of 7-man tablebases. To fully compute them, one will need about 10 PB (10,000 TB) of disk space and 50 TB of RAM. Only the top 10 supercomputers can solve the 8-man problem in 2014. The first 1000-move mate is unlikely to be found until 2020 when a part of a TOP100 supercomputer may be allowed to be used for solving this task.
This information is outdated. https://syzygy-tables.info/ the 7 man Tablebase uses 18 Tbit which is around 0.33bit PER POSITION!
So you today can download and lookup a 7 man TB for around 2k dollars with current nvme ssd prices.
Yes the size of 8 man tablebase will be 110x larger in every domain - the point is that 8 man can be solved in 2022 as hardware has come a long way. Especially on the SSD side you can directly map ssd to gpu memory with directStorage and Cuda can access that.
Im not interested in x-to mate problem. What interests me is the ELO scaling from these tablebases as of today it seems that 6man tb is as good as one weak pruning commit for SF.
For 7 man I lack data.
Leo wrote: ↑Sat Nov 20, 2021 10:59 pm
I found this information a few years ago. Quote:
EGTB
6 man tablebase syzygy {EGTB} 90 GB
7 man tablebase The high speed of generating the tablebases was because of using a supercomputer named Lomonosov (top500). The size of seven-man tablebases is about 140 TB.
Many show interest in what is to expect from 8-man endings. First, take note that the longest 6-man mate took 262 moves (KRN-KNN). Moving to 7-man endings doubled this value. Second, 8-man tablebases include much more endings with both sides having relatively equal strength. All this gives us a strong hope to discover a mate in more than 1000 moves in one of 8-man endgames. Unfortunately the size of 8-man tablebases will be 100 times larger than the size of 7-man tablebases. To fully compute them, one will need about 10 PB (10,000 TB) of disk space and 50 TB of RAM. Only the top 10 supercomputers can solve the 8-man problem in 2014. The first 1000-move mate is unlikely to be found until 2020 when a part of a TOP100 supercomputer may be allowed to be used for solving this task.
This information is outdated. https://syzygy-tables.info/ the 7 men Tablebase uses 18 Tbit which is around 0.33bit PER POSITION!
So you today can download and lookup a 7 man TB for around 2k dollars with current nvme ssd prices.
Yes the size of 8 men tablebase will be 110x larger in every domain - the point is that 8 man can be solved in 2022 as hardware has come a long way. Especially on the SSD side you can directly map ssd to gpu memory with direct Storage and Cuda can access that.
I'm not interested in x-to mate problem. What interests me is the ELO scaling from these tablebases as of today it seems that 6man tb is as good as one weak pruning commit for SF.
For 7 man I lack data.
So how many elo will a chess engine increase by having a 8 men tablebase ? I do not believe that it will increase it by much since most top engines by the time that they have reached 8 men left on the board one of the two engines most of the time has already reached an advantage over the other engine and if NOT their calculating power should let them continue even over an engine that has a 7 men TB.
Geri's checkers is also getting popular, of course he is laughing, this guy play the role of a checkers cheater playing the red stone twice first with one side of the board, and lastly faking a heart attack to turn the checkers board around==>
