Andscacs

Andscacs
by Daniel José Queraltó
http://www.andscacs.com/

Andscacs 0.60j x64
http://www.andscacs.com/andscacs.zip

Andscacs 0.60k x64
- bugfix to avoid using too much CPU when idle
http://www.andscacs.com/andscacs060k.zip

Andscacs 0.62u (x64/w32)

- added information of Nodes per second
- added Static Exchange Evaluation for quiet moves, for depth > 2
- some optimizations for speed gain
- better ordering of moves when the king is in check; try first moving less valuable pieces and add history value of the move
- faster calculation of piece attacks. No need any further of rotated bitboards
- improved some parameters of the evaluation function
- improved values for Late Move Reduction; principal cause of the Elo gain of this version
- cut bad moves/captures in the theoretical last level before Quiesce, if Alpha is raised and there are a minimum of moves already verified
- penalty for advancing pawn that might be weak
- fixed bug on threat evaluation
- improved king safety evaluation
- little hash for evaluations

http://www.andscacs.com/andscacs062u.zip

Andscacs 0.62v (x64/w32)

- Improved some parameters of the evaluation function.
- Improved values for Late Move Reduction.
- New king safety evaluation.

http://www.andscacs.com/andscacs062v.zip

Andscacs 0.62w
http://www.andscacs.com/andscacs062w.zip

http://www.andscacs.com/

Andscacs 0.62x

Changes
* Added knowledge of some basic endings KRBKR, KRNKR, KRKN, KQBKQ, KQNKQ, KBNKR, KBBKR, KNNKR, KBPKB.
* Added some knowledge about endings with bishops of different color.
* Bonus for grouped passed pawns.
* Diminish the evaluation if it’s nearer the 50 move draw rule.
* Augmented velocity of move generation and doing the moves.
* Diminished the possibilities of hash key collision, and fixed the validation of the hash move.
* Optimized a case where it was possible to maintain a move in the hash even if there was none.
* Limited the maximum number of extensions possible before qsearch.
* Fixed bugs in move generation and about storing in history.
* Move ordering:
Favor attacking moves in some cases.
Little penalization for pieces than return to previous square.
Little penalization for king moves when there are many pieces.
* LMR, late move reductions:
Reduce less if it’s an endgame.
Reduce less if it’s in the firsts levels after root moves.
* Penalization for king too open.
* Moves found in hash than permit to return immediately, now are stored as killers and also update the history.
* Optimized the value about material ending that activates some reductions.

Andscacs 0.63

* Prune more moves instead of trying them in some cases.
* If LMR don’t works, also reduces a little in some cases.
* Trapped knight’s evaluation more aggressive.

http://www.andscacs.com/andscacs064.zip

Andscacs 0.64

Changes
1- Some little speed optimizations and bug fixes.
2- Move ordering: At move generation, duplicate value of exchanges of pieces, and increase values of passed pawns moving to ranks 6 or 7.
3- Move generation: start at a1 for white pieces and at h8 for black pieces.
4- Cut more moves at the last alpha_beta before quiescence search.
5- Penalization for trying to win without pawns.
6- Added divide command for doing perft move by move
7- LMR in endgame: increase or decrease it depending on some conditions.
8- Don't do LMR if it's a move of pawn to the rank 7.
9- Optimized hash and solved a bug. Now it's saved also the evaluation of the position.
10- Big optimization of king safety.
11- If the evaluation of one iteration is 0, for the next one clean killers and history.
12- If at the end of quiesce there was no move analyzed, return static evaluation instead of alpha.

Andscacs 0.70 http://www.andscacs.com/andscacs070.zip

http://www.andscacs.com/andscacs071.zip

Andscacs 0.71

Changes:

* Improved evaluation of passed pawns.
* General improvement of speed.
* Fixed some bugs.
* New king safety evaluation.
* Little improvements on LMR.
* Improved mobility values.
* Removed second hash of evaluations. Now uses only main hash for this.

Andscacs 0.74024
http://www.andscacs.com/andscacs074024.zip

[Hide]4th Generation Intel Core i7-4940MX processor (Overclocked to 3.99GHz)
Deep Fritz 12 GUI
1024 mb hash each where possible
3-4-5-6 piece tablebases on Samsung EVO 840 SSD
32 GB RAM (31.9 GB usable)
Ponder off
Perfect_2015s.ctg book (8 moves deep)
40 moves in 18 minutes repeated (Adapted for CCRL)
10 cycles 160 rounds
All engines 64-bit where available[/Hide]
W=25.0 plies; 3,977kN/s; Perfect2015s.ctg  
B=23.1 plies; 8,221kN/s; 21,115 TBAs; Perfect2015s.ctg
[pgn][Event "Andscacs 0.80 64-bit 4CPU Gauntlet-1"]
[Site "Philadelphia"]
[Date "2015.05.17"]
[Round "1.1"]
[White "Andscacs 0.80"]
[Black "Equinox 3.30 x64mp"]
[Result "1-0"]
[ECO "C88"]
[Annotator "0.12;0.11"]
[PlyCount "113"]
[EventDate "2015.05.17"]
[EventType "simul"]
[Source "Russell"]
[TimeControl "40/1080:40/1080:40/1080"]

{Intel(R) Core(TM) i7-4940MX CPU @ 3.10GHz 3999 MHz  W=25.0 plies; 3,977kN/s;
Perfect2015s.ctg  B=23.1 plies; 8,221kN/s; 21,115 TBAs; Perfect2015s.ctg} 1. e4
{B 0} e5 {B 0} 2. Nf3 {B 0} Nc6 {B 0} 3. Bb5 {B 0} a6 {B 0} 4. Ba4 {B 0} Nf6 {
B 0} 5. O-O {B 0} Be7 {B 0} 6. Re1 {B 0} b5 {B 0} 7. Bb3 {B 0} O-O {B 0} 8. h3
{B 0} Bb7 {B 0} 9. d3 {0.12/23 27} d6 {0.11/24 93 (Be7-c5)} 10. a3 {0.10/25 40}
Re8 {0.11/24 14 (Nc6-a5)} 11. Nc3 {0.15/24 48 (Nb1-d2)} Nd4 {0.09/27 46
(Qd8-d7)} 12. Nxd4 {0.28/24 21} exd4 {0.09/25 1} 13. Ne2 {0.25/26 31 (Nc3-d5)}
c5 {0.05/22 9} 14. c3 {0.28/26 20} dxc3 {0.07/22 7} 15. Nxc3 {0.24/24 19} Rc8 {
0.07/23 14 (Bb7-c8)} 16. Qf3 {0.20/23 18} Qd7 {0.07/23 11} 17. Bf4 {0.18/25 18}
Bd8 {0.00/23 71} 18. Rad1 {0.30/25 26 (Ra1-c1)} Ba5 {0.15/25 101 (Bd8-b6)} 19.
Bg5 {0.44/24 18} Bd8 {0.14/24 8} 20. d4 {0.45/24 23} c4 {0.14/23 11} 21. Bc2 {
0.45/26 24} Qc7 {0.14/23 7 (Rc8-b8)} 22. Qf5 {0.53/25 47 (Bg5-f4)} g6 {0.14/27
55 (Re8-e6)} 23. Qf3 {0.53/24 20} Nd7 {0.14/26 28} 24. Bxd8 {0.54/24 63
(Bg5-f4)} Qxd8 {0.12/24 45} 25. Qf4 {0.58/25 30} Qc7 {0.16/23 42 (Re8-e6)} 26.
b4 {0.66/23 20 (Re1-e3)} cxb3 {0.02/24 30 (Nd7-b6)} 27. Bxb3 {0.45/26 23} Nb6 {
0.01/25 9} 28. Re3 {0.55/27 61} Qd7 {0.00/26 39} 29. h4 {0.50/26 48} Nc4 {0.05/
25 13} 30. Rg3 {0.52/25 18 (Bb3xc4)} Rc7 {0.08/24 135} 31. h5 {0.53/23 21} Qe7
{0.24/22 58} 32. Rdd3 {0.88/22 31 (Rg3-g4)} Rd8 {0.25/22 47 (Kg8-h8)} 33. Rdf3
{0.90/24 52 (Rg3-g4)} Kg7 {0.37/22 44 (Kg8-h8)} 34. Rg4 {0.89/26 42} Rdc8 {0.
51/22 45 (Rd8-g8)} 35. Rh3 {1.07/23 41 (Bb3xc4)} Kg8 {0.40/22 7} 36. Ba2 {1.26/
23 49 (Qf4-h6)} Qe6 {0.40/22 17 (Rc8-b8)} 37. Rgg3 {1.66/26 41 (Rg4-h4)} Rd7 {
0.98/22 21} 38. Qh4 {1.79/24 35} Kh8 {0.96/22 12} 39. Bb3 {1.80/24 18 (d4-d5)}
Qe7 {1.05/22 19} 40. Qf4 {1.71/25 23 (h5xg6)} Kg8 {1.07/22 20} 41. Qf5 {2.02/
26 135} Qe8 {1.14/24 38 (d6-d5)} 42. Rh4 {2.18/26 15 (h5xg6)} Re7 {1.14/23 71
(Qe8-e6)} 43. a4 {2.10/25 14 (Qf5-f4)} Qd7 {1.36/22 41 (b5-b4)} 44. hxg6 {2.35/
24 15} fxg6 {1.36/20 0} 45. axb5 {2.12/24 19} axb5 {1.46/22 9 (Re7-g7)} 46.
Nxb5 {2.00/25 25} d5 {1.48/24 70} 47. Qg5 {2.13/25 51 (Qf5xd7)} Rf8 {1.45/21 27
} 48. Nc3 {2.70/23 14} Rg7 {1.82/23 40} 49. e5 {2.81/24 16 (Nc3-a4)} Qe6 {1.68/
23 34 (Rf8-f7)} 50. Qc1 {2.94/24 23 (Rh4-f4)} Rc8 {1.68/22 22 (Qe6-e7)} 51. Bc2
{3.53/26 67 (Qc1-h6)} Rcc7 {1.86/23 32} 52. Qh6 {3.97/24 16 (Rh4-f4)} Rce7 {3.
42/21 80 (Nc4-b2)} 53. Rhg4 {4.10/26 28 (Rh4-h3)} Nxe5 {3.44/22 25 (Qe6-d7)}
54. dxe5 {4.69/27 50} Qxe5 {3.44/21 1} 55. Bb3 {4.70/29 56 (Rg3-e3)} Qd6 {4.86/
24 187} 56. Qd2 {4.96/27 15} Kh8 {4.86/23 12 (Re7-e5)} 57. Bxd5 {5.21/30 29
(Nc3xd5)} 1-0

[/pgn]

[Hide]4th Generation Intel Core i7-4940MX processor (Overclocked to 3.99GHz)
Deep Fritz 12 GUI
1024 mb hash each where possible
3-4-5-6 piece tablebases on Samsung EVO 840 SSD
32 GB RAM (31.9 GB usable)
Ponder off
Perfect_2015s.ctg book (8 moves deep)
40 moves in 18 minutes repeated (Adapted for CCRL)
10 cycles 160 rounds
All engines 64-bit where available[/Hide]
W=21.7 plies; 7,866kN/s; Perfect2015s.ctg
B=28.0 plies; 4,223kN/s; Perfect2015s.ctg
[pgn][Event "Andscacs 0.80 64-bit 4CPU Gauntlet-1"]
[Site "Philadelphia"]
[Date "2015.05.18"]
[Round "2.7"]
[White "Protector 1.6.0 x64"]
[Black "Andscacs 0.80"]
[Result "0-1"]
[ECO "D46"]
[Annotator "0.42;0.22"]
[PlyCount "98"]
[EventDate "2015.05.17"]
[EventType "simul"]
[Source "Russell"]
[TimeControl "40/1080:40/1080:40/1080"]

{Intel(R) Core(TM) i7-4940MX CPU @ 3.10GHz 3999 MHz W=21.7 plies; 7,866kN/s;
Perfect2015s.ctg B=28.0 plies; 4,223kN/s; Perfect2015s.ctg} 1. d4 {B 0} Nf6 {
B 0} 2. Nf3 {B 0} e6 {B 0} 3. c4 {B 0} d5 {B 0} 4. Nc3 {B 0} c6 {B 0} 5. e3 {
B 0} Nbd7 {B 0} 6. Qc2 {B 0} Bd6 {B 0} 7. Bd3 {B 0} O-O {B 0} 8. O-O {B 0} dxc4
{0.22/24 28} 9. Bxc4 {0.42/19 25} e5 {0.16/24 23 (h7-h6)} 10. Bb3 {0.47/20 33
(Bc4-d3)} Qe7 {0.19/21 16} 11. h3 {0.60/20 27} Rb8 {0.19/22 16 (Bd6-c7)} 12. a4
{0.50/18 66 (Bc1-d2)} b6 {0.14/22 19 (a7-a5)} 13. Rd1 {0.74/18 21} Bb7 {0.14/
25 23 (Bd6-b4)} 14. a5 {0.61/18 26 (d4xe5)} e4 {0.14/23 28} 15. Nd2 {0.62/20 19
} Rfe8 {0.13/27 37} 16. axb6 {0.63/20 18} axb6 {0.53/26 29} 17. Ra7 {0.54/21 85
} b5 {0.45/26 21} 18. Ba2 {0.68/21 40} Bb4 {0.39/28 30 (c6-c5)} 19. Bb1 {0.52/
20 30} Bxc3 {0.48/26 28} 20. bxc3 {0.63/22 22} c5 {0.12/26 72} 21. Nb3 {0.53/
22 34} c4 {0.00/27 30} 22. Nc5 {0.54/23 17} Nxc5 {0.00/32 29} 23. Ba3 {0.46/25
35} Qd7 {0.00/30 33 (Qe7-c7)} 24. Bxc5 {0.21/23 43} Ra8 {0.00/30 35} 25. Rxa8 {
0.22/25 27} Rxa8 {0.00/32 23} 26. Re1 {0.16/23 28 (Rd1-f1)} Bc6 {0.00/29 62}
27. Qb2 {0.16/24 27} Qe8 {-0.06/32 21} 28. Bc2 {0.08/24 46 (Kg1-h2)} Nd5 {-0.
06/33 18} 29. Rf1 {0.10/23 19 (Bc2-b1)} f5 {-0.22/30 51} 30. Bb1 {0.00/23 30}
Qg6 {-0.11/27 38 (Qe8-d7)} 31. Qd2 {0.00/25 23} Qh6 {-0.11/27 22 (Qg6-g5)} 32.
Re1 {0.00/25 29 (Kg1-h2)} Qf6 {-0.11/28 36 (Qh6-g6)} 33. Kh2 {-0.14/24 120} Qf7
{-0.11/28 20 (Ra8-e8)} 34. Rc1 {0.00/23 22} Qg6 {-0.12/27 38} 35. Re1 {0.00/25
16} Qe8 {-0.12/28 42 (Qg6-f7)} 36. f3 {0.00/22 15 (Kh2-g1)} Nf6 {-0.43/28 24}
37. f4 {-0.30/22 73} Nd5 {-0.67/28 24 (Qe8-h5)} 38. Rg1 {-0.41/20 16 (Re1-c1)}
Bd7 {-0.98/27 30 (Kg8-h8)} 39. Rc1 {-0.45/22 27} Qh5 {-1.12/29 39 (Kg8-h8)} 40.
Qe1 {-0.57/21 20 (Rc1-g1)} g5 {-1.44/28 24 (Kg8-h8)} 41. Bd6 {-1.29/23 248
(Qe1-g3)} Ra6 {-1.77/27 32 (Qh5-g6)} 42. Be5 {-1.85/20 37} g4 {-2.21/30 86} 43.
h4 {-2.09/21 48 (Qe1-g3)} Qh6 {-2.36/30 32} 44. Rc2 {-2.29/21 33 (Kh2-g1)} Ra3
{-2.74/29 44} 45. Kg1 {-2.28/21 24 (g2-g3)} b4 {-3.14/30 48 (Kg8-f7)} 46. Rb2 {
-3.28/20 93} b3 {-3.75/28 53} 47. h5 {-3.54/20 34 (Kg1-f1)} Qxh5 {-6.32/29 30
(Kg8-f7)} 48. g3 {-5.78/20 25} Ra1 {-7.71/33 48} 49. Rh2 {-6.97/22 34} Qg6 {-8.
07/32 32} 0-1

[/pgn]

[Hide]4th Generation Intel Core i7-4940MX processor (Overclocked to 3.99GHz)
Deep Fritz 12 GUI
1024 mb hash each where possible
3-4-5-6 piece tablebases on Samsung EVO 840 SSD
32 GB RAM (31.9 GB usable)
Ponder off
Perfect_2015s.ctg book (8 moves deep)
40 moves in 18 minutes repeated (Adapted for CCRL)
10 cycles 160 rounds
All engines 64-bit where available[/Hide]
W=20.9 plies; 9,370kN/s; Perfect2015s.ctg  
B=28.5 plies; 4,851kN/s; Perfect2015s.ctg
[pgn][Event "Andscacs 0.80 64-bit 4CPU Gauntlet-1"]
[Site "Philadelphia"]
[Date "2015.05.18"]
[Round "2.13"]
[White "Glaurung 2.2 JA"]
[Black "Andscacs 0.80"]
[Result "0-1"]
[ECO "A29"]
[Annotator "-0.09;-0.11"]
[PlyCount "96"]
[EventDate "2015.05.17"]
[EventType "simul"]
[Source "Russell"]
[TimeControl "40/1080:40/1080:40/1080"]

{Intel(R) Core(TM) i7-4940MX CPU @ 3.10GHz 3999 MHz  W=20.9 plies; 9,370kN/s;
Perfect2015s.ctg  B=28.5 plies; 4,851kN/s; Perfect2015s.ctg} 1. c4 {B 0} e5 {
B 0} 2. Nc3 {B 0} Nf6 {B 0} 3. Nf3 {B 0} Nc6 {B 0} 4. g3 {B 0} d5 {B 0} 5. cxd5
{B 0} Nxd5 {B 0} 6. Bg2 {B 0} Nb6 {B 0} 7. O-O {B 0} Be7 {B 0} 8. d3 {B 0} O-O
{B 0} 9. Be3 {-0.09/19 46} Be6 {-0.11/23 33} 10. Rc1 {0.00/19 41 (d3-d4)} f5 {
-0.21/23 22 (Rf8-e8)} 11. Qd2 {0.07/19 111 (a2-a3)} h6 {-0.13/24 52} 12. Rcd1 {
0.03/19 82 (Qd2-c2)} Qe8 {-0.15/22 26} 13. d4 {-0.09/18 26} Rd8 {-0.27/27 49}
14. Qc1 {-0.13/19 28} e4 {-0.40/25 59} 15. Ne5 {-0.19/20 44} Bb4 {-0.34/24 21
(Nc6xe5)} 16. Nb5 {-0.27/19 71 (f2-f3)} Bxa2 {-0.71/24 34} 17. Nxc7 {-0.45/19
23} Qe7 {-0.76/26 39} 18. Nxc6 {-0.49/19 21} bxc6 {-0.85/27 24} 19. Na6 {-0.56/
20 19} Bc4 {-0.96/25 28} 20. Nxb4 {-0.60/21 31} Qxb4 {-0.90/28 58} 21. Rde1 {
-0.60/21 34} Nd5 {-0.92/25 52 (Qb4-b5)} 22. Bh3 {-0.60/19 17 (Qc1-c2)} Qb5 {-1.
02/27 19} 23. Qc2 {-0.62/19 36 (Qc1-d2)} Rf6 {-1.00/24 24} 24. Bg2 {-1.03/19 17
} a5 {-1.04/25 24} 25. h3 {-1.05/19 40 (Re1-a1)} Nb4 {-1.26/24 28 (c6-c5)} 26.
Qd2 {-0.94/20 26 (Qc2-d1)} c5 {-1.74/27 33} 27. f3 {-1.64/19 35} cxd4 {-1.89/
23 24} 28. fxe4 {-1.54/19 20} Rd7 {-1.92/27 25} 29. Bxd4 {-1.43/20 28 (Be3-f2)}
Nc6 {-2.45/27 34} 30. e3 {-1.31/20 25} Bxf1 {-2.55/29 25} 31. Bxf1 {-1.37/19 26
} Nxd4 {-2.58/28 24 (Qb5-b8)} 32. exd4 {-1.66/19 26} Qb6 {-2.66/28 43} 33. exf5
{-1.68/20 21} Rxd4 {-2.74/29 37} 34. Qf2 {-2.29/19 28} Rb4 {-2.83/30 34 (a5-a4)
} 35. Qxb6 {-2.09/21 26} Rfxb6 {-2.97/33 25} 36. b3 {-2.27/22 33 (Re1-e5)} Kf8
{-3.12/31 24} 37. Ra1 {-3.01/22 26} Rxb3 {-3.25/33 20} 38. g4 {-3.33/23 21}
Rg3+ {-3.23/32 26} 39. Kf2 {-3.35/24 17 (Kg1-h2)} Rc3 {-3.33/34 20} 40. Rxa5 {
-3.54/23 14} Rb2+ {-3.43/35 33} 41. Kg1 {-3.49/26 46} Kf7 {-3.72/32 19
(Rc3-g3+)} 42. Ra6 {-3.15/24 39 (Ra5-a1)} Rg3+ {-3.76/34 19} 43. Kh1 {-3.17/22
3} Rf3 {-3.96/33 35 (Rb2-f2)} 44. Ra1 {-3.58/25 52 (Bf1-c4+)} Kf6 {-4.18/33 25}
45. Kg1 {-3.78/24 42 (Ra1-e1)} Kg5 {-4.41/35 34} 46. Re1 {-3.86/25 45 (Ra1-c1)}
Kh4 {-4.91/32 21 (Kg5-f4)} 47. g5 {-5.52/24 82} Kxg5 {-5.82/32 23 (Rf3-g3+)}
48. Ra1 {-6.66/24 96 (Kg1-h1)} Rg3+ {-6.44/35 29} 0-1

[/pgn]

[Hide]4th Generation Intel Core i7-4940MX processor (Overclocked to 3.99GHz)
Deep Fritz 12 GUI
1024 mb hash each where possible
3-4-5-6 piece tablebases on Samsung EVO 840 SSD
32 GB RAM (31.9 GB usable)
Ponder off
Perfect_2015s.ctg book (8 moves deep)
40 moves in 18 minutes repeated (Adapted for CCRL)
10 cycles 160 rounds
All engines 64-bit where available[/Hide]
Andscacs plays a beautiful pawn storm endgame...defeating Nirvanachess 2.1c
(Move 18 sacrifices the a5 knight favoring a d3 passed pawn ripping apart White's center.)
W=26.6 plies; 4,955kN/s; Perfect2015s.ctg  
B=29.1 plies; 5,010kN/s; Perfect2015s.ctg
[pgn][Event "Andscacs 0.80 64-bit 4CPU Gauntlet-1"]
[Site "Philadelphia"]
[Date "2015.05.19"]
[Round "5.2"]
[White "Nirvanachess 2.1c (popcount)"]
[Black "Andscacs 0.80"]
[Result "0-1"]
[ECO "C88"]
[Annotator "0.00;0.04"]
[PlyCount "102"]
[EventDate "2015.05.17"]
[EventType "simul"]
[Source "Russell"]
[TimeControl "40/1080:40/1080:40/1080"]

{Intel(R) Core(TM) i7-4940MX CPU @ 3.10GHz 3999 MHz  W=26.6 plies; 4,955kN/s;
Perfect2015s.ctg  B=29.1 plies; 5,010kN/s; Perfect2015s.ctg} 1. e4 {B 0} e5 {
B 0} 2. Nf3 {B 0} Nc6 {B 0} 3. Bb5 {B 0} a6 {B 0} 4. Ba4 {B 0} Nf6 {B 0} 5. O-O
{B 0} Be7 {B 0} 6. Re1 {B 0} b5 {B 0} 7. Bb3 {B 0} O-O {B 0} 8. h3 {B 0} Bb7 {
B 0} 9. Nc3 {0.00/26 38} Re8 {0.04/22 23 (d7-d6)} 10. Nd5 {0.00/25 31 (d2-d3)}
Na5 {0.24/22 29} 11. Nxe7+ {0.04/26 35} Qxe7 {0.18/27 48} 12. d3 {0.00/28 33}
c5 {0.10/25 16} 13. Bg5 {0.09/28 67} h6 {0.13/25 25 (Na5xb3)} 14. Bxf6 {0.01/
28 41 (Bg5-h4)} Qxf6 {0.06/23 27} 15. Bd5 {0.04/29 53 (Nf3-d2)} Bxd5 {-0.05/29
25} 16. exd5 {0.00/29 52} Qd6 {-0.11/29 59 (Na5-b7)} 17. c4 {0.00/27 22
(Qd1-d2)} bxc4 {0.00/25 29 (e5-e4)} 18. Qa4 {0.20/26 26} cxd3 {0.00/26 54
(Na5-b7)} 19. Qxa5 {0.00/27 34} Qxd5 {0.00/24 21} 20. Nd2 {0.17/27 48 (b2-b4)}
d6 {0.16/25 24 (Re8-e6)} 21. Qa4 {0.11/25 17 (Ra1-d1)} Reb8 {0.16/25 25} 22.
Qc4 {0.23/26 27} Qd4 {0.07/30 79 (Qd5xc4)} 23. b3 {0.51/27 22} Rb4 {0.00/30 34}
24. Qxd4 {0.42/27 15} Rxd4 {0.23/29 23} 25. Re4 {0.26/27 37 (Ra1-d1)} f5 {0.00/
27 17} 26. Rxd4 {0.54/26 34} cxd4 {0.10/26 32} 27. Rc1 {0.59/26 14} Re8 {0.00/
28 30 (e5-e4)} 28. Kf1 {0.56/26 18 (Nd2-f3)} e4 {0.00/30 20} 29. Ke1 {0.76/26
30} d5 {0.00/28 24} 30. Rc5 {0.56/21 14} Rd8 {-0.25/27 25} 31. b4 {0.58/23 21
(h3-h4)} e3 {0.00/27 33 (f5-f4)} 32. fxe3 {1.23/22 14 (Nd2-f3)} dxe3 {0.00/29
33} 33. Nf3 {1.05/23 9} f4 {0.00/30 32 (d3-d2+)} 34. Ne5 {0.76/24 29} d2+ {-0.
16/33 32} 35. Ke2 {0.76/24 15} g5 {-0.24/28 17 (d5-d4)} 36. Nd3 {0.55/23 52
(Ke2-d1)} Rd6 {-0.69/28 18} 37. Rc7 {0.55/24 23} h5 {-0.75/29 21 (Rd6-e6)} 38.
Rc5 {0.00/23 38 (a2-a4)} Kf7 {-1.00/27 27} 39. a4 {0.00/24 13 (h3-h4)} Kg6 {-1.
10/29 27 (Kf7-e6)} 40. h4 {0.00/25 26} Kf5 {-1.09/33 35} 41. Rc8 {0.00/26 87}
f3+ {-0.74/31 22 (g5xh4)} 42. gxf3 {0.12/28 33} gxh4 {-0.88/32 35} 43. Re8 {0.
00/29 36 (Rc8-f8+)} d4 {-0.28/31 19 (Rd6-g6)} 44. Re5+ {0.00/25 38} Kg6 {-0.42/
33 46} 45. Nf4+ {0.00/28 42} Kf7 {-0.41/33 42} 46. Rxh5 {0.00/29 33 (Re5-f5+)}
Rc6 {-2.03/34 26} 47. Rc5 {0.00/30 28} Rxc5 {-1.84/33 28 (Rc6-f6)} 48. bxc5 {
-2.06/33 24} Ke8 {-1.83/37 19} 49. c6 {-2.44/35 64 (Nf4-h5)} Kd8 {-7.53/35 20
(Ke8-d7)} 50. Ne6+ {-8.25/31 43} Kc8 {-9.07/37 43} 51. Ng5 {-9.36/33 46} Kc7 {
-10.06/36 38} 0-1

[/pgn]

[Hide]4th Generation Intel Core i7-4940MX processor (Overclocked to 3.99GHz)
Deep Fritz 12 GUI
1024 mb hash each where possible
3-4-5-6 piece tablebases on Samsung EVO 840 SSD
32 GB RAM (31.9 GB usable)
Ponder off
Perfect_2015s.ctg book (8 moves deep)
15m/40+15m/40+15m/40 (Adapted for CCRL)
Equation: T minutes / 40 moves repeated, where T = 40 * <elapsed seconds> / 48 = <elapsed seconds> / 1.2
CCRL 40/40: 40 * 18 = 720 = 15 (18 / 1.2)
30 cycles 600 rounds
All engines 64-bit where available[/Hide]
Andscacs 0.81 plays 36.Nb5-c3 evaluated at +1.86 preparing for a surgical strike against cheng4 0.38 demolishing its pawns.
W=29.9 plies; 4,649kN/s; Perfect2015s.ctg  
B=23.8 plies; 8,371kN/s; Perfect2015s.ctg
[pgn][Event "Andscacs 0.81 64-bit 4CPU"]
[Site "Philadelphia"]
[Date "2015.07.19"]
[Round "2.18"]
[White "Andscacs 0.81"]
[Black "cheng4 0.38"]
[Result "1-0"]
[ECO "B92"]
[Annotator "0.15;-0.07"]
[PlyCount "123"]
[EventDate "2015.07.17"]
[EventType "simul"]
[Source "Russell"]
[TimeControl "40/900:20/900:20/900"]

{Intel(R) Core(TM) i7-4940MX CPU @ 3.10GHz 3999 MHz  W=29.9 plies; 4,649kN/s;
Perfect2015s.ctg  B=23.8 plies; 8,371kN/s; Perfect2015s.ctg} 1. e4 {B 0} c5 {
B 0} 2. Nf3 {B 0} d6 {B 0} 3. d4 {B 0} cxd4 {B 0} 4. Nxd4 {B 0} Nf6 {B 0} 5.
Nc3 {B 0} a6 {B 0} 6. Be3 {B 0} e5 {B 0} 7. Nb3 {B 0} Be6 {B 0} 8. f3 {B 0} Be7
{B 0} 9. Be2 {0.15/28 41} Nbd7 {-0.07/22 28} 10. Nd5 {0.06/26 41} Nxd5 {-0.05/
21 24} 11. exd5 {0.00/28 26} Bf5 {-0.08/22 28} 12. Qd2 {0.11/26 25 (0-0)} a5 {
-0.05/21 28 (0-0)} 13. a4 {0.13/25 13} Bh4+ {-0.10/22 25 (Qd8-c7)} 14. g3 {0.
18/27 43} Bf6 {-0.03/23 28} 15. O-O {0.24/24 15} O-O {-0.10/22 27} 16. Bb5 {0.
19/25 29 (Be2-d3)} Qc7 {-0.14/20 23} 17. c4 {0.26/25 15 (Rf1-f2)} b6 {-0.13/20
29 (Bf5-h3)} 18. Qe2 {0.12/22 14} Bh3 {-0.12/21 19 (Bf5-g6)} 19. Rf2 {0.08/24
14 (Rf1-e1)} Be7 {-0.04/20 29} 20. Re1 {0.00/25 14} Nf6 {-0.14/21 20 (f7-f5)}
21. Bc6 {0.39/25 22 (Nb3-d2)} Rab8 {-0.23/22 30} 22. Nd2 {0.25/27 26} h6 {-0.
18/22 28 (Nf6-d7)} 23. Nb1 {0.31/26 15} Qc8 {-0.17/25 27 (Nf6-d7)} 24. Nc3 {0.
56/28 33} Qf5 {0.00/24 30 (Nf6-d7)} 25. b3 {0.47/27 31 (Qe2-c2)} Qg6 {-0.25/23
30} 26. Qc2 {0.46/26 17 (Nc3-b5)} Bf5 {-0.24/23 30 (Qg6xc2)} 27. Qd2 {0.34/25
21} Nd7 {-0.22/22 30} 28. Nb5 {0.56/27 21} h5 {-0.20/21 30 (Bf5-h3)} 29. Rg2 {
0.48/24 21} Bh3 {-0.24/23 30} 30. Rge2 {0.55/23 19 (Bc6xd7)} h4 {-0.18/23 30
(Qg6-f5)} 31. g4 {0.36/29 39} Rfd8 {-0.23/24 22 (Nd7-c5)} 32. Bxd7 {0.49/26 32
(Qd2-c2)} Rxd7 {-0.62/24 31} 33. Qc2 {0.63/30 18 (Kg1-h1)} Qh7 {-0.41/24 31
(Qg6xc2)} 34. Qxh7+ {0.80/30 23} Kxh7 {-0.21/26 26} 35. Ra2 {0.72/29 21
(Be3-c1)} Kg6 {0.00/26 32 (f7-f5)} 36. Nc3 {1.86/33 24} f5 {0.24/26 22} 37. Nd1
{1.86/33 46} fxg4 {0.55/26 34 (Rb8-f8)} 38. Nf2 {1.94/36 26} gxf3 {0.66/26 34
(Rb8-f8)} 39. Nxh3 {2.62/30 39} Kf5 {0.77/25 32} 40. Rf1 {2.89/34 31 (Kg1-f2)}
Kg4 {0.60/26 30} 41. Nf2+ {2.90/37 88} Kf5 {1.08/28 33} 42. Bd2 {3.27/32 30
(Nf2-d1)} h3 {0.70/25 77 (Rb8-f8)} 43. Nh1 {3.67/33 32 (Nf2xh3)} Kg4 {1.30/25
37} 44. Kf2 {3.67/36 28 (Ra2-c2)} Bh4+ {1.90/23 45} 45. Ke3 {3.83/35 87} Rf8 {
1.61/24 52} 46. Nf2+ {3.87/32 23} Kh5 {2.40/24 33} 47. Ne4 {4.00/33 55} f2 {2.
63/24 45} 48. Ke2 {4.25/34 35} g5 {2.81/26 37} 49. Be3 {4.31/33 26} g4 {2.88/
26 46} 50. Rb2 {4.03/29 31 (Ra2-a1)} Rg7 {2.93/23 37} 51. Rbb1 {4.66/31 49} Rg6
{3.38/23 63 (Rf8-f4)} 52. Bxb6 {4.99/34 54} Rf4 {3.61/25 44} 53. Nxf2 {5.18/32
36} g3 {4.00/26 89} 54. hxg3 {5.23/34 52} Rxg3 {4.23/26 22} 55. Be3 {5.31/33
56 (Rf1-h1)} Rff3 {3.25/25 47} 56. Bd2 {5.06/32 37} Rf5 {3.46/25 28 (Rf3-f8)}
57. Ne4 {6.21/33 52} Rg2+ {3.96/26 42} 58. Kd3 {6.36/33 32} Rxf1 {4.13/25 42}
59. Rxf1 {6.70/32 32} Kg4 {4.47/23 21} 60. Nxd6 {7.19/33 57 (Rf1-h1)} h2 {5.19/
22 63} 61. Rh1 {7.56/35 81} Bg3 {5.62/23 61 (Kg4-h3)} 62. Ne4 {15.11/31 73
(c4-c5)} 1-0

[/pgn]

Andscacs 0.82

From the author:

Is not much stronger, maybe 10 elo or a little more, but I decided to do it anyway as I'm starting to rewrite big parts of it and I wanted to publish my latest version.

http://andscacs.com/andscacs082.zip

It's stronger than 10 elo in multiprocessor mode, because I solved some bugs and also improved it when is using more than 16 cores. The previous version was weaker with 32 cores than with 16, and this version is maybe 20 elo stronger with 32 vs 16 cores.

Changes:
* Solved little bugs.
* Some small speed improvements.
* Improvements on multi processor mode.
* Improved razoring, null move and lmr.
* Don't evaluate king safety on some simple endgames.

I suppose a new stronger version will not be available for some months. It's not easy to reach 3000 CCRL elo starting from 0

Andscacs 0.82
- new version has several silent updates where version number has remained unchanged
http://andscacs.com/andscacs082.zip

Andscacs 0.86
http://www.andscacs.com/andscacs086.zip

WarMaster wrote:

Andscacs 0.86134
http://www.andscacs.com/

This is a bugfix release, right? or there is an elo improvement??

good ... but I think there was an improvement ... but nobody said anything yet ... so I'm not sure ... hug ...

Ahmed Kamal wrote:

WarMaster wrote:

Andscacs 0.86134
http://www.andscacs.com/

This is a bugfix release, right? or there is an elo improvement??

Not sure, but it's not an official release as such.