Equisetum - UCI chess engine written in C++11, by Rhys Rustad-Elliott and Litov Alexander
Rating Chess Engines Diary CEDR=3498
Equisetum uses Fail-Hard AB-search and NNUE evaluation. This approach resulted in an engine of moderate strength. One of the few strong engines using a fail-hard approach
Equisetum uses an NNUE (efficiently updatable neural network) as an evaluation function.
## Strength
Current Equisetum is estimated to be somewhere before Stockfish 11 and Stockfish 12 strength.
Equisetum 1.0 vs other engines:
Obsidian 11.0 | 5/18 | -8 | 18 Games |
Rebel 16.3 | 4/15 | -7 | 15 Games |
Ethereal 14.25 | 4/12 | -4 | 12 Games |
Chess-System-Tal-2 | 4/12 | -4 | 12 Games |
Uralochka 3.41 dev3 | 4/12 | -4 | 12 Games |
Seer 2.8.0 | 3/12 | -6 | 12 Games |
RubiChess 20240112 | 2.5/12 | -7 | 12 Games |
Caissa 1.17 | 1.5/12 | -9 | 12 Games |
Orion 1.0 | 9.5/10 | +9 | 10 Games |
Jet 1.2 | 7/10 | +4 | 10 Games |
Lizard 10.3 | 3/10 | -4 | 10 Games |
Berserk 20240311 | 2.5/10 | -5 | 10 Games |
Clover 6.1.19 | 3.5/9 | -2 | 9 Games |
Belette 2.0.0 | 8/8 | +8 | 8 Games |
Viridithas 12.0.0 | 3/8 | -2 | 8 Games |
Black Marlin 9.0 | 3/8 | -2 | 8 Games |
Alexandria 6.1.0 | 1.5/8 | -5 | 8 Games |
Stockfish 16.1 | 1.5/8 | -5 | 8 Games |
RofChade 3.1 | 2/7 | -3 | 7 Games |
Velvet 7.0.0 | 4/6 | +2 | 6 Games |
Deep HIARCS 15.3 | 4/6 | +2 | 6 Games |
Jim Ablett compiles:
Comments
Post a Comment