TalkChess.com

It seems today's net 1209 is the best with 7 +/- 6 Elo points in 8000 games over the past best net. Score of Sergio_1209 vs Sergio_29.07_0109: 1180 - 1073 - 1747 [0.513] 4000 ... Sergio_NNUE playing White: 900 - 266 - 834 [0.658] 2000 ... Sergio_NNUE playing Black: 280 - 807 - 913 [0.368] 2000 ... ...

NNUE and regular stockfish are the same in correspondence time control or analysis I find them completely different, worlds apart. You can always guide the position to places where NNUE outperforms regular Stockfish (just give bonuses for positions where they disagree, as the disagreement grows you...

My one great mistake in doing a build was cheap RAM. Never again. If there is one key place to save money, this isn't it. What ram did you put in it? When I bought my ram I had little choice. All of the kits I wanted were out of stock and I had already waited more than two weeks trying to find "goo...

That is pretty much what Gigabyte support said, and I tried all that, and no luck so far. --Jon If it were me I would reseat my CPU repeatedly. I did it by the book the first time and it didn't work. It took several tries to get it right. Doing it on a flat surface is a must. Flexing the MB is a NO...

NNUE plays beautiful positional chess at LTC (10min+20sec). Sample games Yes, but does it play stronger is the question. Beauty is very subjective. ELO isn't. NNUE and regular stockfish are the same in correspondence time control or analysis NNUE has been around long enough for this to be determine...

My build is not going so well. I am using a TRX40 AORUS (Gigabyte) MASTER mb with a 3970x. Put it all together, no POST. The board has the RAM status light red and the LEDs have code 60, which is apparently just before the "NVRAM initialization" step. The RAM modules are CORSAIR CMK32GX4M2A2400C14,...

As even with bucket sizes of 4 there would virtually always be something of very little value in the bucket, I guess this would be determined by the amount of pressure there is on the TT. If there is little or no TT pressure I agree there will be plenty of space in the TT. But this isn't the situat...

My "optimal in this respect" only referred to the strategy of sticking to the (cache-line-sized) bucket the initial hashing function maps you to, and replace what you consider least valuable there, rather than trying to look elsewhere to see if you can find something even less valuable to replace. ...

Rough maths suggests he has at least 1000 cores to do that. If we call several = 5 5 billion fens a day = 50000 fens per second, which if he samples at 25% implies 12500 games per second. One game every 0.08 ms. For massively fast bullet that needs 1000 cores at least, I think. 1000 cores would all...

Very recently ttak added an arch that supports switching based on pieces left rather than gameply. Having weights for each piece might be even better. But the best would be to have a small net trained to select the best evaluation net. What I've seen is that NNUE is extremely sensitive to opening s...