Supernova 2.4

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

Dark Knight
Posts: 8
Joined: Mon Aug 31, 2020 1:03 am
Full name: Minkai Yang

Supernova 2.4

Post by Dark Knight »

Hi all!

I'm glad to announce that Supernova 2.4 is released. :D
https://github.com/MichaeltheCoder7/Sup ... s/tag/v2.4

Changes:

Added lazy SMP. Supernova can now use more than one CPU when thinking!
Fixed a Zobrist hashing bug regarding en passant.
Fixed an issue in static exchange evaluation.
Various optimizations.
Changed data type in the 8x8 board.
Encoded castling rights.
Made quit command more stable.
Compiled with GCC 11.2.

Testing Results:

Supernova 2.4 (4 CPUs) vs Supernova 2.3:
Time control: 10+0.1
Games: 910
Elo difference: 152.2 +/- 18.1

Supernova 2.4 (1 CPU) vs Supernova 2.3:
Time control: 10+0.1
Games: 3000
Elo difference: 17.2 +/- 9.1

Please let me know if you found any issues or bugs.
User avatar
CMCanavessi
Posts: 1142
Joined: Thu Dec 28, 2017 4:06 pm
Location: Argentina

Re: Supernova 2.4

Post by CMCanavessi »

Plans for FRC support?
Follow my tournament and some Leela gauntlets live at http://twitch.tv/ccls
Dark Knight
Posts: 8
Joined: Mon Aug 31, 2020 1:03 am
Full name: Minkai Yang

Re: Supernova 2.4

Post by Dark Knight »

CMCanavessi wrote: Tue Sep 14, 2021 11:25 pm Plans for FRC support?
Not planning for that at the moment. It might be complicated since I need another evaluation for FRC.
User avatar
Rebel
Posts: 6991
Joined: Thu Aug 18, 2011 12:04 pm

Re: Supernova 2.4

Post by Rebel »

Gauntlet Supernova 2.4

Games : 1000

Elo Pool : 2710

Progress can be followed at - http://rebel13.nl/b/grl.htm

Page refreshes itself every 30 seconds.
90% of coding is debugging, the other 10% is writing bugs.
User avatar
Rebel
Posts: 6991
Joined: Thu Aug 18, 2011 12:04 pm

Re: Supernova 2.4

Post by Rebel »

Supernova 2.4 : -7 elo

Within the error bars.

Details in the archive - https://prodeo.actieforum.com/t358p25-g ... sults#4497
90% of coding is debugging, the other 10% is writing bugs.
Dark Knight
Posts: 8
Joined: Mon Aug 31, 2020 1:03 am
Full name: Minkai Yang

Re: Supernova 2.4

Post by Dark Knight »

Rebel wrote: Wed Sep 15, 2021 4:14 pm Supernova 2.4 : -7 elo

Within the error bars.

Details in the archive - https://prodeo.actieforum.com/t358p25-g ... sults#4497
Thanks for testing. Unfortunately, with 1 CPU, elo gain is not much so it needs a lot more games to be accurate. It's probably more worthwhile to test with 4 CPUs.
Chessqueen
Posts: 5582
Joined: Wed Sep 05, 2018 2:16 am
Location: Moving
Full name: Jorge Picado

Re: Supernova 2.4

Post by Chessqueen »

Dark Knight wrote: Wed Sep 15, 2021 7:53 pm
Rebel wrote: Wed Sep 15, 2021 4:14 pm Supernova 2.4 : -7 elo

Within the error bars.

Details in the archive - https://prodeo.actieforum.com/t358p25-g ... sults#4497
Thanks for testing. Unfortunately, with 1 CPU, elo gain is not much so it needs a lot more games to be accurate. It's probably more worthwhile to test with 4 CPUs.
Here is an interesting game between similar rated programs Shredder8 is from 2003, but is still a great program

[pgn][Event ", Blitz:10'+5""]
[Site "?"]
[Date "2021.09.15"]
[Round "1"]
[White "Supernova 2.4"]
[Black "Shredder 8"]
[Result "0-1"]
[ECO "D78"]
[WhiteElo "2750"]
[BlackElo "2800"]
[Annotator "0.38;0.00"]
[PlyCount "98"]
[TimeControl "600+5"]

{W=38.5 ply; 796kN/s B=15.8 ply; 6,590kN/s} 1. d4 {[%emt 0:00:00]} Nf6 {
[%emt 0:00:00]} 2. Nf3 {[%emt 0:00:00]} d5 {[%emt 0:00:00]} 3. g3 {
[%emt 0:00:00]} g6 {[%emt 0:00:00]} 4. Bg2 {[%emt 0:00:00]} Bg7 {[%emt 0:00:00]
} 5. O-O {[%emt 0:00:00]} O-O {[%emt 0:00:00]} 6. c4 {[%emt 0:00:00]} dxc4 {
[%emt 0:00:00]} 7. Nbd2 {[%emt 0:00:00]} c6 {[%emt 0:00:00]} 8. Nxc4 {
[%emt 0:00:00]} Be6 {[%emt 0:00:00]} 9. Qc2 {Both last book move [%emt 0:00:00]
} Qd5 {[%eval 0,17] [%emt 0:00:26]} 10. b3 {(Nce5) [%eval 38,24] [%emt 0:00:31]
} Bf5 {(Nbd7) [%eval 26,17] [%emt 0:00:38]} 11. Qd1 {
(Qd2) [%eval 32,24] [%emt 0:00:30]} Be4 {[%eval 19,18] [%emt 0:00:32]} 12. Ncd2
{[%eval 35,25] [%emt 0:00:29]} Bf5 {(Bxf3) [%eval 33,18] [%emt 0:00:27]} 13. e3
{(Bb2) [%eval 32,23] [%emt 0:00:28]} Nbd7 {(Be6) [%eval 28,17] [%emt 0:00:30]}
14. Ba3 {(Nh4) [%eval 29,25] [%emt 0:00:27]} Rfe8 {[%eval 40,17] [%emt 0:00:22]
} 15. Nh4 {[%eval 28,25] [%emt 0:00:22]} Be4 {[%eval 53,18] [%emt 0:00:25]} 16.
Nxe4 {[%eval 31,25] [%emt 0:00:22]} Nxe4 {[%eval 51,19] [%emt 0:00:23]} 17. f3
{(Qc2) [%eval 27,26] [%emt 0:00:21]} Nd6 {[%eval 8,16] [%emt 0:00:19]} 18. f4 {
[%eval 22,25] [%emt 0:00:20]} Qa5 {[%eval 19,17] [%emt 0:00:25]} 19. Bc1 {
(Bb2) [%eval 16,24] [%emt 0:00:20]} c5 {(e6) [%eval -15,16] [%emt 0:00:17]} 20.
Bb2 {[%eval 0,21] [%emt 0:00:19]} cxd4 {[%eval -2,15] [%emt 0:00:16]} 21. Bxd4
{[%eval 13,21] [%emt 0:00:02]} Bf6 {[%eval -5,17] [%emt 0:00:25]} 22. Qe1 {
(Rc1) [%eval 12,23] [%emt 0:00:19]} Qa6 {(Qb5) [%eval -9,15] [%emt 0:00:09]}
23. Qb4 {[%eval 11,24] [%emt 0:00:19]} Rac8 {
(Bxd4) [%eval -8,16] [%emt 0:00:22]} 24. Bxf6 {
(Rf2) [%eval 38,26] [%emt 0:00:18]} exf6 {[%eval -27,17] [%emt 0:00:18]} 25.
Rfd1 {(Qd4) [%eval 13,27] [%emt 0:00:18]} Nc5 {[%eval -4,18] [%emt 0:00:26]}
26. Bf1 {[%eval 13,28] [%emt 0:00:17]} Qc6 {[%eval 2,17] [%emt 0:00:21]} 27.
Rac1 {[%eval 13,26] [%emt 0:00:17]} Nde4 {[%eval 0,17] [%emt 0:00:17]} 28. Qd4
{(Bc4) [%eval 0,21] [%emt 0:00:16]} a5 {[%eval -21,17] [%emt 0:00:23]} 29. f5 {
(Bc4) [%eval 14,25] [%emt 0:00:16]} g5 {[%eval -29,16] [%emt 0:00:12]} 30. Ng2
{[%eval -1,26] [%emt 0:00:16]} Qc7 {[%eval -34,17] [%emt 0:00:16]} 31. Qc4 {
(Qb2) [%eval 5,25] [%emt 0:00:15]} b5 {(Re5) [%eval -68,17] [%emt 0:00:34]} 32.
Qxb5 {[%eval -12,28] [%emt 0:00:15]} Qe5 {[%eval -6,15] [%emt 0:00:06]} 33. Bc4
{[%eval -25,29] [%emt 0:00:15]} Qb2 {[%eval -5,16] [%emt 0:00:14]} 34. Qxa5 {
[%eval -25,30] [%emt 0:00:14]} Nd3 {(Qf2+) [%eval -81,16] [%emt 0:00:24]} 35.
Bxf7+ {(Rxd3) [%eval -25,24] [%emt 0:00:14]} Kg7 {
(Kf8) [%eval -423,14] [%emt 0:00:15]} 36. Rc7 {
(Bc4) [%eval -305,21] [%emt 0:00:14]} Qf2+ {
(Rxc7) [%eval -555,15] [%emt 0:00:15]} 37. Kh1 {[%eval -411,24] [%emt 0:00:24]}
Kh6 {[%eval -578,16] [%emt 0:00:14]} 38. Bh5 {
(Bc4) [%eval -456,23] [%emt 0:00:13]} Kxh5 {[%eval -707,15] [%emt 0:00:22]} 39.
h3 {[%eval -609,23] [%emt 0:00:23]} Kh6 {[%eval -1261,15] [%emt 0:00:24]} 40.
Qa7 {(Rxd3) [%eval -775,24] [%emt 0:00:22]} Rxc7 {
(Qxf5) [%eval -32615,16] [%emt 0:00:21]} 41. Qxc7 {
(Qb7) [%eval -32758,26] [%emt 0:00:11]} Nxg3+ {[%eval -32617,15] [%emt 0:00:05]
} 42. Kh2 {(Qxg3) [%eval -32759,34] [%emt 0:00:11]} Ne2 {
(Ne5) [%eval -32617,16] [%emt 0:00:20]} 43. Qd6 {
[%eval -32760,45] [%emt 0:00:11]} Ne5 {[%eval -32760,13] [%emt 0:00:00]} 44.
Qxf6+ {[%eval -32761,100] [%emt 0:00:02]} Kh5 {[%eval -32761,11] [%emt 0:00:00]
} 45. Qxg5+ {[%eval -32762,100] [%emt 0:00:00]} Kxg5 {
[%eval -32762,7] [%emt 0:00:00]} 46. h4+ {[%eval -32763,100] [%emt 0:00:00]}
Kg4 {[%eval -32763,7] [%emt 0:00:00]} 47. Rd4+ {
[%eval -32764,100] [%emt 0:00:00]} Nxd4 {[%eval -32764,5] [%emt 0:00:00]} 48.
exd4 {[%eval -32765,100] [%emt 0:00:00]} Kf3 {
(Nf3+) [%eval -32765,3] [%emt 0:00:00]} 49. dxe5 {
[%eval -32766,100] [%emt 0:00:00]} Qxg2# {[%eval -32766,1] [%emt 0:00:00]} 0-1[/pgn]
Do NOT worry and be happy, we all live a short life :roll: