TalkChess.com

Dann Corbit wrote: ↑

Tue Aug 21, 2018 11:32 pm

I just bought a 1080 Ti because you can get them under $700 now.

will it overheat?

PLS I HAVE TO KNOW

thanks a lot.

syzygy wrote: ↑

Tue Aug 21, 2018 2:14 pm

Generating 8-piece tables with the same algorithm would require a machine with 64 TB RAM.

One such machine would be difficult to get. Could that be divided into a cluster?

Milos wrote: ↑

Mon Aug 20, 2018 11:52 pm

That crappy PR stunt preprint was obviously never accepted for publication anywhere. It was just uploaded to arxiv.org.

I wish I had time and hardware to write crappy preprints like that one

Michel wrote: ↑

Tue Aug 21, 2018 4:31 am

Recall that RdM is a mathematician. A syzygy is in fact a specific (somewhat advanced) mathematical concept:

think he's retired.. but might still want to write a paper on the subject

Dann Corbit wrote: ↑

Mon Aug 20, 2018 5:41 pm

I hope that this effort will become very famous. It is really something incredible.

dunno about fame, but syzygy site will sure ramp up its visits

it will eventually saturate the 20x256 NN.. might saturate at a higher ELO with a bigger net, but that's not clear

as of now saturation can be checked for with some statistical time-series test, but just from looking at it, I doubt it reached that point.

ponder is an option, not a joke

think it's in the rules that tweaks can be made within season if a play-affecting issue is found.

Karol Majewski wrote: ↑

Mon Aug 13, 2018 12:05 pm

If Ruy Lopez and Petroff give black 0.40 then why people keep playing the king's gambit 2.f4? Again, after 2.f4 white is much worse but is he lost? We don't know it yet.

We play the king's gambit because it's fun. The damiano doesn't seem like fun at all. More like masochism