Since you’re here...

We hope you will consider supporting us today. We need your support to continue to exist, because good entries are more and more work time. Every reader contribution, however big or small, is so valuable. Support "Chess Engines Diary" even a small amount– and it only takes a minute. Thank you.
============================== My email: jotes@go2.pl



Minke 5.0.0 - new version chess engine

on


Minke - UCI chess engine
Github:https://github.com/enfmarinho/Minke
Author: Eduardo Marinho

Working in progress UCI compatible chess engine done in order to learn and to get fun. Based on NNUE evaluation, Bitboard board representation and a lot of search enhancements.

Minke 5.0.0 what's new?

Results against Minke 4.0.0

STC:

Elo   | 117.25 +- 16.13 (95%)
SPRT  | 8.0+0.08s Threads=1 Hash=32MB
LLR   | 2.27 (-2.89, 2.25) [0.00, 3.00]
Games | N: 658 W: 278 L: 64 D: 316
Penta | [0, 20, 111, 162, 36]

https://eduardomarinho.dev/test/148/

LTC:

Elo   | 128.36 +- 14.90 (95%)
SPRT  | 40.0+0.40s Threads=1 Hash=128MB
LLR   | 2.26 (-2.89, 2.25) [0.00, 3.00]
Games | N: 628 W: 268 L: 46 D: 314
Penta | [0, 12, 93, 184, 25]

https://eduardomarinho.dev/test/151/

Opening book used for testings was UHO_Lichess_4852_v1.epd

Changes

  • Improved Aspiration Windows #45 #60
  • Implemented entry prefetching and aligned memory allocation for the TT #48 #56
  • "depth-preferred" replacement strategy for the TT #49
  • Improved usage of the TT #50 #51 #52 #61 #65
  • Refined Late Move Reductions #53 #57 #58
  • Improved cache locality during NNUE Accumulator updates #54
  • Reduced TT entry keys to 16 bits #55
  • Integrated ProbCut into the Negamax search for earlier branch pruning #59
  • Introduced continuation history heuristics for 1 and 2 plies to improve move ordering #62
  • Bug fixes #44 #46 #47 #63 #64 #66 #67


Comments

Post a Comment