Worlds most powerful supercomputer.

[Leo]

"The computer, called Summit, is housed at the department's Oak Ridge National Laboratory in Oak Ridge, Tenn. Summit's performance can top out at 200,000 trillion calculations per second, or 200 petaflops, the lab said in a release Friday.
The Massachusetts Institute of Technology's Technology Review reported the computer is 60 percent faster than the supercomputer that was previously the world's most powerful. That computer is called the Sunway TaihuLight and is housed in Wuxi, China."

It cost $200,000,000. It also will cost a lot of money to run and maintain it. I personally would have used the money to make a payment on the national debt.

[Lion]

Will all of this be useless vs AI such as Leela or Alpha Google?

[Vinvin]

https://www.top500.org/news/summit-up-a ... plication/

[Cardoso]

Very nice, for years I've been reading news about colossal machines with astronomical computing power that cost millions of dollars to built and a big amount also to maintain and run. Now, how about putting this machine to really good use like finding the cure for cancer? If mankind knows so much about the human genome, and knows how to built these incredible machines, maybe it's time to tackle this cancer problem. I know it's not anything near interesting from the economic point of view to find the cure for cancer, and it's much more profitable to keep selling the current drugs for this disease. But since this all comes from tax payers, how about public opinion to start talking about this? How about knowledge people completely independent from big pharma companies, start using this astronomical power to really help mankind? Isn't cancer a big enough problem to try to solve? I red in the article "the system’s peak rating for deep learning performance is something on the order of 3.3 exaflops", can't this immense power be used to find the cure for cancer and other families destroying diseases?

[syzygy]

Will all of this be useless vs AI such as Leela or Alpha Google?

With all the GPUs in that system, yes.

[duncan]

Will all of this be useless vs AI such as Leela or Alpha Google?

With all the GPUs in that system, yes.

Would the supercomputer be useful for you in constructing tablebases and if so and you had it for a year, what would you be able to do with it. ?

[Leo]

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.

[syzygy]

Will all of this be useless vs AI such as Leela or Alpha Google?

With all the GPUs in that system, yes.

Would the supercomputer be useful for you in constructing tablebases and if so and you had it for a year, what would you be able to do with it. ?

My generator doesn't benefit from GPUs or clusters, so I would have to write a new one and in my spare time that might take more than a year. (And it's not clear to me that GPUs are useful for generating TBs in a distributed manner.)

[duncan]

With all the GPUs in that system, yes.

Would the supercomputer be useful for you in constructing tablebases and if so and you had it for a year, what would you be able to do with it. ?

My generator doesn't benefit from GPUs or clusters, so I would have to write a new one and in my spare time that might take more than a year. (And it's not clear to me that GPUs are useful for generating TBs in a distributed manner.)

I thought it also has over 9,000 cpus
https://www.top500.org/news/summit-up-a ... plication/

The IBM-built system is comprised of 4,608 nodes, each one housing two Power9 CPUs and six NVIDIA Tesla V100 GPUs.

[Dann Corbit]

Very nice, for years I've been reading news about colossal machines with astronomical computing power that cost millions of dollars to built and a big amount also to maintain and run. Now, how about putting this machine to really good use like finding the cure for cancer? If mankind knows so much about the human genome, and knows how to built these incredible machines, maybe it's time to tackle this cancer problem. I know it's not anything near interesting from the economic point of view to find the cure for cancer, and it's much more profitable to keep selling the current drugs for this disease. But since this all comes from tax payers, how about public opinion to start talking about this? How about knowledge people completely independent from big pharma companies, start using this astronomical power to really help mankind? Isn't cancer a big enough problem to try to solve? I red in the article "the system’s peak rating for deep learning performance is something on the order of 3.3 exaflops", can't this immense power be used to find the cure for cancer and other families destroying diseases?

After analyzing the human genome, they discovered that what they really needed to understand was protein folding.
Protein folding is really, really hard.