Commercial Release of Ethereal 13.00 (NNUE) for AVX2 Systems

[AndrewGrant]

Good news!
With the right net Ethereal 13.00 commercial could be strongly improved
Ethereal 12.75 sse3 github with Zeus.net is already stronger than Fat Fritz 2.0 (Chessbase, Stockfish 13 FF 2.1 github version is much stronger) and Dragon 2.0. ​

Please Andrew, unlock NNUE on the free version, to let testers to experiment. If you like I can give you all the games 12.75 NNUE games + Zeus.net (used also by AI-15) to study why with alien nets Ethereal become so strong, and rebuild your own net.
I think that testing 2 versions of the same engine is not enough, the performance must be calculated considering all the best engines availables.

PS: @Stefan Waiting for your SPCC accurate ranking with the "alien nets" with 6000 games, if possible
Now switching from Arena 3.5.1 to Banksiagui 0.46 that is more stable on engine matches.



Regards, AlexChess
My SuperBlitz always updated: https://www.chess.com/forum/view/game-a ... e-fritz-17

Please stop making these posts. You only serve to confuse people with them.

"With the right net Ethereal 13.00 commercial could be strongly improved ", this is technically true, but when you say it a reader will take it to mean that they can use a different net, one from Stockfish, in Ethereal. This is not the case. It will not work. They are not compatible. The architecture is different. The quantization is different. The weight layouts are different.

I am not "unlocking" NNUE on the free version. It is why there is a commercial version. This would be like asking the Komodo team to "unlock" NNUE by back porting it to the K12 version that is free.

I have absolutely no interest in the results of Ethereal 12.75 using Stockfish networks. The entire point of making that version (Etherlito, a jab at Houdini and Fire), was to PROVE how trivial it was to socket in a Stockfish network and make your engine appear as if it knows how to play chess, when really you've done nothing but replace the majority of it.

[AlexChess]

OK sorry, I'll post them only on my page. I was thinking to do a useful task as fine tuning and bugs reporting tester

It's a big frustration that I cannot test anymore last Ethereal 13.00 version, in my personal challenge to improve it with fine tuning.

PS: Since it is allowed, I hope that someone will fork the 13.00 github GPLv3 version compiling it also for Windows 10 ARM64 (since you say it's trivial) and adding SF NNUE support again.

Regards,
AlexChess

[Chessqueen]

OK sorry, I'll post them only on my page. I was thinking to do a useful task as fine tuning and bugs reporting tester

It's a big frustration that I cannot test anymore last Ethereal 13.00 version, in my personal challenge to improve it with fine tuning.

PS: Since it is allowed, I hope that someone will fork the 13.00 github GPLv3 version compiling it also for Windows 10 ARM64 (since you say it's trivial) and adding SF NNUE support again.

Regards,
AlexChess

My congratulations to Ethereal for securing the spot the it earned from the very beginning at TCEC After the first 4 games it secured the last place and with great effort it did NOT let any other Engine take the place away, it kept it until the very end. Well Ethereal almost did the same as Kasparov

[CMCanavessi]

My congratulations to Ethereal for securing the spot the it earned from the very beginning at TCEC After the first 4 games it secured the last place and with great effort it did NOT let any other Engine take the place away, it kept it until the very end. Well Ethereal almost did the same as Kasparov

It's still DivP... counting CPU engines, only 4 are ahead of Ethereal. All the rest are below.

[Chessqueen]

My congratulations to Ethereal for securing the spot the it earned from the very beginning at TCEC After the first 4 games it secured the last place and with great effort it did NOT let any other Engine take the place away, it kept it until the very end. Well Ethereal almost did the same as Kasparov

It's still DivP... counting CPU engines, only 4 are ahead of Ethereal. All the rest are below.

Next time it will probably be on top once Ethereal get dropped to a lower division, so it will eventually climb back to div P again

[AlexChess]

Don't be rude, I have enjoyed Ethereal very much until 13.00

My disappoint is related to the fact that, also buying the commercial version, it would not work fine on my Mac mini M1 and the github version is a dead end, just a demo weaker than 12.75 + NNUE. I consider NNUEs like Syzygy tablebases. Every programmer adds them to his engine, also if they are identical for all, or endgames play is much weaker. Without NNUE today an engine is <void>. Every programmer starts from an existing engine, adding then his own algorithms (better if the engine is hybrid) and Stockfish Team is more than happy if you share your knowledge forking SF and releasing your modified code.

Restarting all from scratch is at least to be considered stoic

Regards, AlexChess

[mclane]

Some results of ethereal 13.00 (NNUE) so far:

https://prodeo.actieforum.com/t445-stoc ... 1000-games

[AlexChess]

Hi CMCanavessi,

have you solved all the crashes experienced using Ethereal 13.00?

Regards, AlexChess

[AlexChess]

Hi CMCanavessi,

Ethereal 13 standard github 1CPU loses vs Stockfish 14 and derivatives although it is 3 times faster on my test

Could you add also the no-NNUE version to your engine tournament?

Regards, AlexChess

[mclane]

Ethereal is stronger then stockfish 13/14 without NNUE.