aryan1508 / bit-genie

UCI chess engine in C++

Bit-Genie

Bit-Genie is an open-source chess engine written in C++ by Aryan Parekh. It has support for the UCI protocol.

About

TLDR; I got addicted to chess programming

With all the lockdowns in 2020, I decided that it would be a good time to start working on a new project. I have lots of love for programming; I have lots of love for chess, so a chess software seemed like a good start. I ended up creating a small & simple chess GUI which although wasn't easy, didn't satisfy me. I thought it would be interesting if I could write a program that could play good chess and possibly beat me. And that's how I started writing Bit-Genie. What I didn't know was chess programming is very addictive , and it would turn out to be more than a small project. What inspired me the most was the huge community of open-source chess engines like Ethereal

Elo progression

The estimated strength over the different versions

CCRL Blitz (2'+1")

Version	Elo
Bit-Genie 1	1771
Bit-Genie 2	1879
Bit-Genie 4	2230
Bit-Genie 5	2400
Bit-Genie 7	2727
Bit-Genie 8	2860
Bit-Genie 9	3067

Bit-Genie on the CCRL

Rebel's gambit rating list 40/2

Version	Elo
Bit-Genie 5	2450
Bit-Genie 6	2620
Bit-Genie 7	2760
Bit-Genie 8	2870
Bit-Genie 9	3050

Rebel's 40/2 rating list

---

Acknowledgements

The development wouldn't have been possible if not for the huge resources available online like the Chess programming Wiki. The works of the following people have influenced the progress of Bit-Genie greatly

• Special thanks to Kim Kåhre and Finn Eggers for guiding me during the neural network's development
• Jay Honnold
• Terje Kirstihagen
• Andrew Grant
• Kieren Pearson
• Alexander Litov

bit-genie's Issues

Please add ARMv9.2, SVE2, Dot Product and Neon support

It's sad that it's not possible to compile the newest version on Android and Linux devices.

Bit-Genie v9 loses on time in CuteChess

i successfully compiled your newest v9 source code into a 64-bit Linux binary .. then, playing in CuteChess, the engine loses on time at the very last move of the timecontrol : i gave 8 minutes per 40 moves plus 3 seconds bonus (and 1000 ms margin) .. this happens with both White and Black .. see screenshots below .. the engine seems to calculate exactly 15 sec each move but fails to mention the remaining time ..

is some counter 1 off ?
then it should be an easy fix !

Bit-Genie draws endgames with mate scores

Hi Aryan,

Frank Quisinsky reported problems with endgame mates, when resigning/ajudicating is not set.
http://talkchess.com/forum3/viewtopic.php?f=2&t=78885&start=20#p922429

It looks like a different problem than that you seem to have fixed in PR#208.
#208

He also added a zip file with pgn examples.
Note that he is using the Shredder GUI and I have no idea if he has logs too.
(This GUI has a problem with converting certain high scores, but that's more a cosmetic
problem and leads to non standard mate scores in PGN)
As you can see this was Bit-Genie 9.19 I had compiled a while ago.
Here is the first game from it (you can see it cannot convert K+Q vs. K to a win):

[Event "40 Moves in 20 min"]
[Site "FCP Tourney-KI, FQ-3"]
[Date "2022.02.12"]
[Round "7.33"]
[White "Bit-Genie 9.19 dev"]
[Black "Black Marlin 4.0 NN"]
[Result "1/2-1/2"]

1. e4 {book 0} e6 {book 0}
2. d4 {book 0} d5 {book 0}
3. Nc3 {book 0} Nf6 {book 0}
4. Bg5 {book 0} Be7 {book 0}
5. e5 {book 0} Nfd7 {book 0}
6. Bxe7 {book 0} Qxe7 {book 0}
7. Qd2 {book 0} O-O {book 0}
8. f4 {book 0} c5 {book 0}
9. dxc5 {book 0} Nc6 {book 0}
10. Nf3 {+0.53/31 38} Nxc5 {-0.10/24 42}
11. Qe3 {+0.57/34 38} a6 {-0.10/27 40}
12. a3 {+0.67/37 38} Nd7 {+0.07/26 1:08m}
13. Bd3 {+0.74/34 38} Qc5 {+0.06/25 34}
14. Qf2 {+0.74/32 38} h6 {+0.07/25 1:04m}
15. O-O {+0.87/33 38} Qxf2+ {+0.20/23 27}
16. Kxf2 {+0.84/35 38} f6 {0.00/22 27}
17. exf6 {+0.86/33 38} Rxf6 {+0.17/25 25}
18. Ke3 {+1.13/33 38} Nc5 {+0.26/24 25}
19. Be2 {+1.16/36 38} b6 {+0.18/25 1:25m}
20. g3 {+1.24/34 38} Bb7 {+0.07/24 53}
21. b4 {+1.32/36 38} Nd7 {+0.36/23 26}
22. Kd2 {+1.41/36 38} Raf8 {+0.34/27 1:07m}
23. Rae1 {+1.41/34 38} Bc8 {+0.27/24 26}
24. Kc1 {+1.45/34 38} b5 {+0.35/24 25}
25. h4 {+1.39/35 38} Nb6 {+0.43/24 35}
26. Bd3 {+1.60/34 38} Bd7 {+0.48/26 20}
27. Ne5 {+1.70/38 38} Nxe5 {+0.52/28 55}
28. Rxe5 {+1.70/40 38} Kf7 {+0.57/28 48}
29. Rfe1 {+1.49/41 38} g6 {+0.64/28 1:00m}
30. Ne2 {+1.86/44 38} Ke7 {+0.68/26 16}
31. Nd4 {+1.86/44 38} Kd6 {+0.74/25 23}
32. Kd2 {+1.86/42 38} Rg8 {+0.73/22 24}
33. Nb3 {+1.85/38 38} Na4 {+0.72/26 23}
34. R1e3 {+1.86/40 38} Rc8 {+0.53/25 21}
35. Re2 {+1.86/37 38} Be8 {+0.74/25 41}
36. Nd4 {+1.86/41 38} Bd7 {+0.78/28 23}
37. Re1 {+1.80/42 38} Rc3 {+0.77/28 18}
38. Ra1 {+1.81/40 38} Rc7 {+0.71/28 38}
39. Nf3 {+1.90/39 38} Nb2 {+0.67/22 22}
40. Be2 {+1.90/36 38} a5 {+0.72/24 33}
41. Rb1 {+2.10/40 29} Nc4+ {+0.77/23 24}
42. Bxc4 {+2.15/43 29} bxc4 {+0.79/25 19}
43. b5 {+1.89/43 29} Rc5 {+0.91/28 39}
44. Nd4 {+2.15/39 29} Rc8 {+1.07/27 18}
45. Rbe1 {+2.13/42 29} Rc7 {+0.91/30 57}
46. Ke3 {+1.84/39 29} Rc5 {+0.91/29 38}
47. Rb1 {+2.06/40 29} Rc8 {+0.91/27 22}
48. a4 {+2.48/47 29} Kc5 {+0.91/28 16}
49. c3 {+2.48/49 29} Rb8 {+0.91/27 22}
50. Kd2 {+2.49/45 29} Rb6 {+0.91/25 19}
51. Rbe1 {+2.54/47 29} Kd6 {+1.36/32 56}
52. Ke3 {+2.54/46 29} Rb8 {+1.36/31 22}
53. Nf3 {+2.90/49 29} Kc5 {+1.36/30 14}
54. Nd2 {+2.90/50 29} Kd6 {+1.36/30 15}
55. Rd1 {+2.90/52 29} Kc7 {+1.36/30 16}
56. Kd4 {+2.90/50 29} Kd6 {+1.36/30 15}
57. Nf3 {+2.90/48 29} Rbf8 {+1.36/29 17}
58. Rde1 {+2.90/46 29} Bc8 {+1.36/26 19}
59. Rh1 {+2.90/47 29} Bd7 {+1.36/30 48}
60. Rg1 {+2.90/47 29} Re8 {+1.36/30 16}
61. Nh2 {+2.96/43 29} Rff8 {+1.36/30 23}
62. Ng4 {+3.35/50 29} h5 {+1.36/28 19}
63. Ne3 {+3.39/56 30} Re7 {+1.72/30 28}
64. Rg5 {+3.39/55 29} Be8 {+1.72/31 18}
65. Re1 {+3.39/56 30} Ref7 {+1.72/31 17}
66. Nc2 {+3.39/54 30} Rf5 {+1.72/29 34}
67. Re5 {+3.39/60 30} Rxg5 {+1.61/31 27}
68. Rxg5 {+3.53/57 30} Rf5 {+1.60/32 17}
69. Ne1 {+3.53/59 30} Bf7 {+1.59/33 1:19m}
70. Nf3 {+3.53/58 30} Be8 {+2.10/29 45}
71. Nd2 {+3.53/57 30} Bf7 {+1.59/34 31}
72. Nf1 {+3.53/62 30} Be8 {+1.59/34 17}
73. Ne3 {+3.53/61 30} Rf8 {+1.59/34 16}
74. Re5 {+3.53/60 30} Bf7 {+1.59/34 17}
75. Nf1 {+3.53/61 30} Rb8 {+1.59/33 24}
76. Nh2 {+3.53/62 30} Rb7 {+1.59/34 26}
77. Nf3 {+3.53/56 30} Be8 {+1.59/35 33}
78. Rg5 {+3.54/55 30} Rf7 {+1.66/28 33}
79. Re5 {+3.53/54 30} Rb7 {+1.59/35 29}
80. Rg5 {+3.54/55 30} Rg7 {+1.59/33 54}
81. Ne5 {+3.71/41 29} Rg8 {+1.79/28 18}
82. Ke3 {+3.85/40 29} Kc5 {+1.85/27 22}
83. g4 {+3.79/41 29} Rh8 {+2.12/27 1:07m}
84. Nxg6 {+3.61/40 29} Bxg6 {+2.27/27 20}
85. Rxg6 {+3.91/36 29} hxg4 {+2.32/27 37}
86. Rxe6 {+3.98/37 29} d4+ {+1.96/26 51}
87. cxd4+ {+2.53/33 29} Kb4 {+1.92/28 22}
88. f5 {+3.05/35 29} Kxa4 {+1.81/27 44}
89. b6 {+3.32/37 29} Kb5 {+2.17/27 48}
90. f6 {+3.55/37 29} Ka6 {+2.27/27 26}
91. d5 {+3.72/36 29} a4 {+2.22/26 19}
92. Kf4 {+3.89/35 29} g3 {+2.15/27 19}
93. Kxg3 {+4.60/32 29} Rf8 {+2.71/29 1:10m}
94. d6 {+6.29/32 29} a3 {+3.73/25 1:30m}
95. d7 {+6.85/33 29} a2 {+4.73/25 1:43m}
96. Re1 {+7.17/31 29} Kb7 {+5.15/26 5:31m}
97. Kg4 {+8.04/32 29} Kc6 {+5.69/24 21}
98. Ra1 {+9.25/35 29} Kxd7 {+5.78/24 11}
99. Kf5 {+9.42/34 29} Kc6 {+6.17/26 1:01m}
100. Rxa2 {+10.14/34 29} Kxb6 {+6.18/24 18}
101. h5 {+11.04/31 29} Rh8 {+6.27/22 13}
102. Kg6 {+11.70/30 29} Rg8+ {+6.58/22 8}
103. Kh7 {+12.51/34 29} Rc8 {+6.63/22 11}
104. Kg7 {+12.90/33 30} c3 {+7.13/22 10}
105. f7 {+14.03/30 30} c2 {+7.65/20 18}
106. Rxc2 {+18.32/30 30} Rxc2 {+7.33/19 12}
107. f8=Q {+19.76/28 30} Rc7+ {+8.73/19 20}
108. Kg6 {+19.82/27 30} Rc6+ {+7.35/19 9}
109. Kg5 {+21.61/26 30} Rc5+ {+6.43/19 10}
110. Qxc5+ {+M-952/27 30} Kxc5 {+M-1351/48 15}
111. Kf5 {+14.11/31 30} Kd4 {+M-1351/48 4}
112. h6 {+14.10/27 30} Ke3 {+M-1351/53 3}
113. Ke6 {+14.21/29 30} Kf3 {+M14/86 13}
114. Ke7 {+14.27/29 30} Kf2 {+M15/124 4}
115. Kf6 {+14.21/31 30} Kf3 {+M15/113 4}
116. Ke5 {+14.22/32 30} Ke3 {+M13/89 10}
117. Kd5 {+14.21/31 30} Kf3 {+M12/84 8}
118. h7 {+M-959/25 30} Ke3 {+M11/91 4}
119. h8=Q {+M-959/39 30} Kf4 {+M10/255 5}
120. Qd4+ {+M-959/38 30} Kf3 {+M9/255 2}
121. Qe4+ {+M-959/37 29} Kg3 {+M9/255 6}
122. Kc4 {+M-959/53 29} Kh3 {+M6/255 10}
123. Qe6+ {+M-959/44 29} Kh4 {+M8/255 2}
124. Qd6 {+M-959/53 29} Kg5 {+M9/255 0}
125. Qd2+ {+M-959/52 29} Kf5 {+M11/255 1}
126. Qf2+ {+M-959/56 29} Ke5 {+M11/255 2}
127. Qe3+ {+M-959/57 29} Kf6 {+M9/255 0}
128. Qf3+ {+M-959/59 29} Ke5 {+M10/255 1}
129. Qg3+ {+M-959/64 17} Ke6 {+M10/255 0}
130. Qg8+ {+M-959/64 10} Ke7 {+M10/255 0}
131. Qg5+ {+M-959/64 0} Ke6 {+M9/255 0}
132. Qg4+ {+M-959/64 0} Ke7 {+M7/255 2}
133. Qg7+ {+M-959/64 1} Ke6 {+M8/255 0}
134. Qg6+ {+M-959/64 0} Ke5 {+M8/255 0}
135. Qe8+ {+M-959/64 0} Kf4 {+M10/255 0}
136. Qb8+ {+M-959/64 0} Kf3 {+M9/255 2}
137. Qb7+ {+M-959/64 0} Kf4 {+M9/255 0}
138. Qd7 {+M-959/64 0} Kg3 {+M9/255 0}
139. Qf5 {+M-959/64 0} Kh4 {+M8/255 2}
140. Qf6+ {+M-959/64 0} Kg3 {+M8/255 0}
141. Qg6+ {+13.09/64 0} Kf4 {+M7/255 0}
142. Qg8 {+M-959/64 0} Kf5 {+M9/255 0}
143. Kd5 {0.00/64 0} Kf4 {+M9/255 0}
144. Qg2 {0.00/64 0} Ke3 {+M7/255 0}
145. Qe4+ {0.00/64 0} Kf2 {+M6/255 0}
146. Qc2+ {0.00/64 0} Kf3 {+M7/255 1}
147. Qf5+ {0.00/64 0} Ke3 {+M7/255 1}
148. Qe4+ {0.00/64 0} Kf2 {+M6/255 0}
149. Qc2+ {0.00/64 0} Kf3 {+M7/255 0}
150. Qf5+ {0.00/64 0} Ke2 {+M7/255 0}
151. Qg4+ {0.00/64 0} Kd2 {+M7/255 1}
152. Qf4+ {0.00/64 0} Kc3 {+M8/255 0}
153. Qc4+ {0.00/64 0} Kb2 {+M7/255 0}
154. Qb4+ {0.00/64 0} Ka2 {+M5/255 1}
155. Qd2+ {0.00/64 0} Ka1 {+M4/255 0}
156. Qa5+ {0.00/64 0} Kb2 {+M6/255 0}
157. Qb6+ {0.00/64 0} Kc3 {+M7/255 0}
158. Qd4+ {0.00/64 0} Kc2 {+M6/255 0}
159. Qc4+ {0.00/64 0} Kd1 {+M6/255 1}
160. Qd3+ {0.00/64 0} Ke1 {+M6/255 0}
161. Qg3+ {0.00/64 0} Kd2 {+M7/255 0}
162. Qf4+ {0.00/64 0} Kd1 {+M6/255 49}
163. Qa4+ {0.00/64 0} Ke2 {0.00/255 3}
164. Qg4+ {0.00/64 0} Ke1 {0.00/255 0}
165. Qe4+ {0.00/64 0} Kf2 {0.00/255 0}
1/2-1/2

A serious bug!

As of March 2020, Bit-Genie isn't a top-tier chess engine, but we're getting somewhere.

Seriously? update that, you are already there ;)

Incomplete PV during search

PV collected during search is often not displayed completely when printing out the info string

go startpos, 2 mb hash

go
info depth 1 seldepth 1 nodes 43 score cp 64 time 1 pv b1c3
info depth 2 seldepth 2 nodes 106 score cp 0 time 2 pv b1c3 b8c6
info depth 3 seldepth 3 nodes 254 score cp 59 time 2 pv b1c3 b8c6 g1f3
info depth 4 seldepth 6 nodes 624 score cp 7 time 3 pv d2d4 b8c6 e2e4 g8f6
info depth 5 seldepth 9 nodes 2749 score cp 54 time 4 pv b1c3 b8c6 g1f3 g8f6 d2d4
info depth 6 seldepth 10 nodes 6297 score cp 18 time 6 pv b1c3 b8c6 d2d4 g8f6 e2e4 e7e6
info depth 7 seldepth 12 nodes 8686 score cp 40 time 8 pv b1c3 b8c6 d2d4 g8f6 g1f3 d7d5 e2e3
info depth 8 seldepth 13 nodes 15588 score cp 20 time 11 pv b1c3 b8c6 e2e4 g8f6 d2d4 d7d5 e4d5 f6d5
info depth 9 seldepth 20 nodes 43019 score cp 33 time 19 pv b1c3 g8f6 g1f3 d7d5 d2d4 b8c6 e2e3 e7e6 h1g1
info depth 10 seldepth 20 nodes 179201 score cp 13 time 49 pv b1c3
info depth 11 seldepth 20 nodes 133483 score cp 33 time 71 pv b1c3 d7d5 d2d4 g8f6 g1f3 e7e6 c1f4 f8d6 e2e3 e8g8
info depth 12 seldepth 24 nodes 1323096 score cp 26 time 300 pv e2e4
info depth 13 seldepth 24 nodes 516927 score cp 26 time 386 pv e2e4 e7e5
info depth 14 seldepth 24 nodes 1515767 score cp 26 time 633 pv e2e4 e7e5
info depth 15 seldepth 27 nodes 6286785 score cp 38 time 1731 pv e2e4 e7e5 b1c3

Ponder and FRC

Hi! Do you have plans to implement ponder and/or chess 960? We really need more chess 960 engines.

UCI score display

Hi Aryan,

I discovered your nice little and quite new chess program a few days ago.
In a previous compilation I noticed the missing time management, but you have added this now.

Today I compiled it again and it seems it is ready now for running vs. other programs,
it just has a littlle glitch in displaying the eval for GUIs, instead of a 'cp' value with three digits,
(cp actually means centipawn) it just shows one digit always

Thus when it has an eval between 0.00 and 0.99 it will always print cp 0 instead of something between 0 and 99
and if it has an eval of e.g. 2.00 and 2.99 it will always show cp 2. The GUI will interpret cp 2 as 0.02.

Examples:

>Bit-Genie_r107-64(1): go wtime 90478 btime 108859 movestogo 30
<Bit-Genie_r107-64(1): info depth 1 seldepth 4 nodes 65 score cp 0 time 34 pv f3e2 
<Bit-Genie_r107-64(1): info depth 2 seldepth 4 nodes 166 score cp 0 time 34 pv f3e2 f8c5 
<Bit-Genie_r107-64(1): info depth 3 seldepth 7 nodes 2976 score cp 0 time 35 pv f3f4 f8b4 d4e5 
<Bit-Genie_r107-64(1): info depth 4 seldepth 7 nodes 2946 score cp 0 time 37 pv f3f4 e8f6 d4e5 f8d6 
<Bit-Genie_r107-64(1): info depth 5 seldepth 10 nodes 38699 score cp 0 time 49 pv f3g4 e8d6 f5f6 g7f6 c4c5 
<Bit-Genie_r107-64(1): info depth 6 seldepth 10 nodes 93667 score cp 0 time 75 pv f3f4 e8d6 f5f6 g7f6 c4c5 d6c4 
<Bit-Genie_r107-64(1): info depth 7 seldepth 12 nodes 330986 score cp 0 time 163 pv f3e2 e8d6 f5f6 g7f6 e2d3 d6e8 d3e3 

>Bit-Genie_r107-64(1): go wtime 51778 btime 80347 movestogo 17
<Bit-Genie_r107-64(1): info depth 1 seldepth 1 nodes 13 score cp -2 time 33 pv c2b1 
<Bit-Genie_r107-64(1): info depth 2 seldepth 3 nodes 57 score cp -2 time 34 pv c2b1 c4c3 
<Bit-Genie_r107-64(1): info depth 3 seldepth 3 nodes 148 score cp -2 time 35 pv c2b1 c4c3 b1a1 
<Bit-Genie_r107-64(1): info depth 4 seldepth 5 nodes 834 score cp -2 time 35 pv c2b1 c4c3 b1a1 c3c2 
<Bit-Genie_r107-64(1): info depth 5 seldepth 7 nodes 4484 score cp -3 time 37 pv c2c1 e4c3 c1c2 c3a2 c2b1 
<Bit-Genie_r107-64(1): info depth 6 seldepth 8 nodes 31631 score cp -4 time 45 pv c2b1 c4d3 b1c1 e4c3 h2h3 c3a2 
<Bit-Genie_r107-64(1): info depth 7 seldepth 10 nodes 18790 score cp -4 time 50 pv c2b1 c4d3 b1c1 e4c3 h2h3 c3e2 c1b1 

Here is its first win vs. new releases Rustic Alpha 2

[Event "Test@CuteCGdv"]
[Site "RWBC-CAPPUCCINO Win7U64 Q8200 2.33Ghz + Nvidia GT 710"]
[Date "2021.03.19"]
[Round "3"]
[White "Rustic_Alpha2-64"]
[Black "Bit-Genie_r107-64"]
[Result "0-1"]
[ECO "A40"]
[GameDuration "00:03:29"]
[GameEndTime "2021-03-19T12:01:13.857 Mitteleuropäische Zeit"]
[GameStartTime "2021-03-19T11:57:43.896 Mitteleuropäische Zeit"]
[Opening "Queen's pawn"]
[PlyCount "76"]
[TimeControl "40/120"]

1. d4 {book} c6 {book} 2. Nf3 {book} d5 {book} 3. e3 {book} Bg4 {book}
4. Be2 {+0.40/7 1.92s} Nd7 {0.00/7 3.19s} 5. O-O {+0.55/7 1.94s}
e6 {0.00/7 3.19s} 6. Nc3 {+0.45/7 1.96s} Ngf6 {0.00/7 3.19s}
7. Ne5 {+0.30/7 1.69s} Bf5 {0.00/7 3.20s} 8. g4 {+0.70/6 2.02s}
Bg6 {0.00/7 3.20s} 9. Nxg6 {+0.75/7 2.04s} hxg6 {0.00/7 3.20s}
10. g5 {+0.65/7 2.07s} Ne4 {+0.01/7 3.20s} 11. f4 {+0.30/7 2.10s}
Nxg5 {+0.01/7 3.20s} 12. fxg5 {-0.35/7 2.12s} Qxg5+ {+0.01/7 3.21s}
13. Kf2 {-0.40/6 2.16s} Rxh2+ {+0.03/7 3.21s} 14. Ke1 {-0.35/6 2.19s}
Qh4+ {+0.03/7 3.21s} 15. Kd2 {-0.25/6 2.23s} e5 {+0.03/7 3.21s}
16. dxe5 {-0.35/7 2.27s} d4 {+0.03/7 3.21s} 17. exd4 {-0.50/6 2.30s}
Qxd4+ {+0.03/7 3.21s} 18. Ke1 {-3.70/6 2.35s} Qh4+ {+0.04/7 3.22s}
19. Kd2 {-3.80/6 2.39s} Nxe5 {+0.06/7 3.22s} 20. Nd5 {-4.87/6 1.86s}
Rd8 {+0.07/7 3.22s} 21. Kc3 {-6.40/7 2.46s} Rxd5 {+0.09/7 3.22s}
22. Rf4 {-8.20/7 2.56s} Rxd1 {+0.09/7 3.23s} 23. Rxh4 {-7.40/8 2.62s}
Rxe2 {+0.09/8 3.23s} 24. b3 {-7.27/7 2.68s} f5 {+0.09/7 3.23s}
25. Rh8 {-7.00/8 2.70s} Ree1 {+0.09/8 3.24s} 26. Kb2 {-7.10/8 2.83s}
f4 {+0.10/8 3.24s} 27. Rh7 {-7.70/7 2.91s} f3 {+0.11/8 3.24s}
28. Rh2 {-9.45/7 3.00s} Re2 {+0.12/8 3.25s} 29. Be3 {-12.55/8 3.10s}
Rxa1 {+0.15/8 3.25s} 30. Rxe2 {-13.50/9 2.87s} fxe2 {+0.16/9 3.26s}
31. Bd2 {-13.75/9 2.47s} Rd1 {+0.16/8 3.26s} 32. Bc3 {-18.30/8 3.58s}
Nf3 {+0.21/8 3.27s} 33. b4 {-19.25/8 3.77s} Nd2 {+0.23/8 3.27s}
34. Bxd2 {-21.70/9 3.99s} Rxd2 {+0.24/8 3.28s} 35. b5 {-M8/9 3.89s}
e1=Q {+M7/8 3.29s} 36. a4 {-M6/7 4.65s} Qd1 {+M5/7 3.30s} 37. bxc6 {-M4/7 5.12s}
Rxc2+ {+M3/7 3.31s} 38. Kb3 {-M2/6 5.81s} Qd3# {+M1/8 3.33s, Black mates} 0-1

Do you plan to make a pre-release, or would you mind, if I add it to my XB/UCI chronology?
https://docs.google.com/spreadsheets/d/1GXoy3lGhdKPJbatdw2_2dhWT3TkpthPcNtHs51MQydM

Guenther

Support for larger hash sizes?

I downloaded Bit genie 9 compiled for windows to play against some engines. I tried to set the hash to 10 GB but the gui limited the hash to 3 GB. However, it would not initialize until I set it to 1.5 GB. If you are competing in TCEC, some engines use 256 GB. Not sure if you are interested in that but the engine not initializing is the issue.