HP Enterprise unveils The Machine, with 160 TB of memory

[cdani]

Some big tablebases at sight!

https://venturebeat.com/2017/05/16/hp-e ... e-machine/

[Dann Corbit]

40 nodes.
I am guessing this thing costs some serious money

[Nordlandia]

40 nodes.
I am guessing this thing costs some serious money

8-man lomonosov tablebases not considered earlier than 2018.

http://rybkaforum.net/cgi-bin/rybkaforu ... pid=562705

[Leo]

40 nodes.
I am guessing this thing costs some serious money

8-man lomonosov tablebases not considered earlier than 2018.

http://rybkaforum.net/cgi-bin/rybkaforu ... pid=562705

"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."

[Nordlandia]

"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."

It's 5 years since 7-men were completed. They might consider generate 8-man next year.

[duncan]

"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."

It's 5 years since 7-men were completed. They might consider generate 8-man next year.

any idea what elo gain it would give a computer?

[CheckersGuy]

"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."

It's 5 years since 7-men were completed. They might consider generate 8-man next year.

any idea what elo gain it would give a computer?

I dont think we even know how much we would gain from 7 piece db.

[Nordlandia]

8-man is 25% of the whole set.

Is it accurate to say Chess is solved by 1/ 4 ?

[Dann Corbit]

8-man is 25% of the whole set.

Is it accurate to say Chess is solved by 1/ 4 ?

Each set is exponentially larger.

At the very most one part in 72 quadrillion solved, but probably a lot smaller portion than that.

Plus the fact that there are no 8 man file sets available.

[mhull]

Some big tablebases at sight!

https://venturebeat.com/2017/05/16/hp-e ... e-machine/

This is just a prototype that uses SDRAM to simulate the proposed memory fabric. Production systems will ship with hundreds of petabytes installed, out-ot-the-box using memrister technology for the byte-addressable memory fabric (or similar, non-volatile, low/zero power, memory technology).

large systems will ship with thousands of petabytes and hundreds of compute nodes, each with tens or hundreds of processors.

The goal is to be able to process vast datasets in realtime (national/world facial recognition in realtime, for instance), enabling decision support on a scale not possible heretofor. Applications undreamt of will arise.

This is considered a "disruptive technology". When the first units ship, current super computers will be as obsolete as a Commodore 64.