Latest Crafty-25.1-ACCA

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

Moderator: Ras

User avatar
MikeB
Posts: 4889
Joined: Thu Mar 09, 2006 6:34 am
Location: Pen Argyl, Pennsylvania

Latest Crafty-25.1-ACCA

Post by MikeB »

This is the version playing in the 2016 ACCA Championship with Syzygy TB support (Nalimov is no longer supported)

[d]8/k7/3p4/p2P1p2/P2P1P2/8/8/K7 w - -

Code: Select all

Crafty v25.1-ACCA

machine has 24 processors
( this was set to one core with Fine #70 st =1 second)
White(1): st 1
search time set to 1.00.
White(1): 
White(1): g
        time surplus   0.00  time limit 1.00 (1.00)
        depth     time       score   variation (1)
          1     0.00/1.00     0.98   1. Kb2                        
       ...
          7->   0.00/1.00     1.18   1. Kb2 Kb6 2. Kc3 Kb7 3. Kd3 Kb6 4. Kc4
          8     0.00/1.00     1.14   1. Kb2 Kb6 2. Kc3 Kb7 3. Kd3 Ka6 4. Ke3
                                     Kb6
          8->   0.00/1.00     1.14   1. Kb2 Kb6 2. Kc3 Kb7 3. Kd3 Ka6 4. Ke3
                                     Kb6
          9     0.00/1.00     0.98   1. Kb2 Kb6 2. Kc3 Kb7 3. Kd3 Ka6 4. Ke3
                                     Kb6 5. Kf3
          9     0.00/1.00     0.98   1. Kb1 Ka6 2. Kc2 Kb6 3. Kd2 Ka7 4. Ke3
                                     Kb6 5. Kf3
          9->   0.00/1.00     0.98   1. Kb1 Ka6 2. Kc2 Kb6 3. Kd2 Ka7 4. Ke3
                                     Kb6 5. Kf3
        ...
         25     0.01/1.00     3.19   1. Kb1 Kb6 2. Kc2 Kc7 3. Kd3 Kb6 4. Ke3
                                     Kc7 5. Kf2 Kd8 6. Kg3 Ke7 7. Kh4 Kf7
                                     8. Kg5 Kg7 9. Kxf5 Kf7 10. Kg5 Kg7 11. f5
                                     Kf7 12. f6 Ke8 13. Kg6
         25->   0.01/1.00     3.19   1. Kb1 Kb6 2. Kc2 Kc7 3. Kd3 Kb6 4. Ke3
                                     Kc7 5. Kf2 Kd8 6. Kg3 Ke7 7. Kh4 Kf7
                                     8. Kg5 Kg7 9. Kxf5 Kf7 10. Kg5 Kg7 11. f5
                                     Kf7 12. f6 Ke8 13. Kg6
         26     0.01/1.00     3.35   1. Kb1 Kb6 2. Kc2 Kc7 3. Kd3 Kb6 4. Ke3
                                     Kc7 5. Kf2 Kd8 6. Kg3 Ke7 7. Kh4 Kf7
                                     8. Kg5 Kg7 9. Kxf5 Kf7 10. Kg5 Kg7 11. f5
                                     Kf7 12. f6 Ke8 13. Kg6 Kf8
         26->   0.01/1.00     3.35   1. Kb1 Kb6 2. Kc2 Kc7 3. Kd3 Kb6 4. Ke3
                                     Kc7 5. Kf2 Kd8 6. Kg3 Ke7 7. Kh4 Kf7
                                     8. Kg5 Kg7 9. Kxf5 Kf7 10. Kg5 Kg7 11. f5
                                     Kf7 12. f6 Ke8 13. Kg6 Kf8
         27     0.01/1.00     3.25   1. Kb1 Kb6 2. Kc2 Kc7 3. Kd3 Kb6 4. Ke3
                                     Kc7 5. Kf2 Kd8 6. Kg3 Ke7 7. Kh4 Kf7
                                     8. Kg5 Kg7 9. Kxf5 Kf7 10. Kg5 Kg7 11. f5
                                     Kf7 12. f6 Ke8 13. Kg6 Kf8 14. f7
         27->   0.01/1.00     3.25   1. Kb1 Kb6 2. Kc2 Kc7 3. Kd3 Kb6 4. Ke3
                                     Kc7 5. Kf2 Kd8 6. Kg3 Ke7 7. Kh4 Kf7
                                     8. Kg5 Kg7 9. Kxf5 Kf7 10. Kg5 Kg7 11. f5
                                     Kf7 12. f6 Ke8 13. Kg6 Kf8 14. f7
         28     0.02/1.00     3.53   1. Kb1 Kb6 2. Kc2 Kc7 3. Kd3 Kb6 4. Ke3
                                     Kc7 5. Kf2 Kd8 6. Kg3 Ke7 7. Kh4 Kf7
                                     8. Kg5 Kg7 9. Kxf5 Kf7 10. Kg5 Kg7 11. f5
                                     Kf7 12. f6 Ke8 13. Kg6 Kf8 14. f7 Ke7
         ...
         36->   0.11/1.00     4.99   1. Kb1 Kb6 2. Kc2 Kc7 3. Kd3 Kb6 4. Ke3
                                     Kc7 5. Kf2 Kd8 6. Kg3 Ke7 7. Kh4 Kf7
                                     8. Kg5 Kg7 9. Kxf5 Kf7 10. Kg5 Kg7 11. f5
                                     Kf7 12. f6 Kf8 13. Kg4 Ke8 14. Kf4 Kf8
                                     15. Kg5 Kf7 16. Kf5 Kf8 17. Ke6 Ke8
                                     18. Kxd6 Kf7
         37     0.26/1.00     Won    1. Kb1 Kb6 2. Kc2 Kc7 3. Kd3 Kb6 4. Ke3
                                     Kc7 5. Kf2 Kd8 6. Kg3 Ke7 7. Kh4 Kf7
                                     8. Kg5 Kg7 9. Kxf5 Kf7 10. Ke4 Kg6 11. f5+
                                     Kf6 12. Kd3 Kf7 13. f6 Kg6 14. f7 Kxf7
                                     15. Kc4 Kf8 16. Kb5 Ke7 17. Kxa5 <EGTB>
         37->   0.26/1.00     Won    1. Kb1 Kb6 2. Kc2 Kc7 3. Kd3 Kb6 4. Ke3
                                     Kc7 5. Kf2 Kd8 6. Kg3 Ke7 7. Kh4 Kf7
                                     8. Kg5 Kg7 9. Kxf5 Kf7 10. Ke4 Kg6 11. f5+
                                     Kf6 12. Kd3 Kf7 13. f6 Kg6 14. f7 Kxf7
                                     15. Kc4 Kf8 16. Kb5 Ke7 17. Kxa5 <EGTB>
         ...        
         50     0.77/1.00     Won    1. Kb1 Kb6 2. Kc2 Kc7 3. Kd3 Kb6 4. Ke3
                                     Kc7 5. Kf2 Kd8 6. Kg3 Ke7 7. Kh4 Kf7
                                     8. Kg5 Kg7 9. Kxf5 Kf7 10. Ke4 Kg6 11. f5+
                                     Kf6 12. Kd3 Kf7 13. f6 Kg6 14. f7 Kxf7
                                     15. Kc4 Kf8 16. Kb5 Ke7 17. Kxa5 <EGTB>
         50->   0.77/1.00     Won    1. Kb1 Kb6 2. Kc2 Kc7 3. Kd3 Kb6 4. Ke3
                                     Kc7 5. Kf2 Kd8 6. Kg3 Ke7 7. Kh4 Kf7
                                     8. Kg5 Kg7 9. Kxf5 Kf7 10. Ke4 Kg6 11. f5+
                                     Kf6 12. Kd3 Kf7 13. f6 Kg6 14. f7 Kxf7
                                     15. Kc4 Kf8 16. Kb5 Ke7 17. Kxa5 <EGTB>
        time=1.01(100%)  nodes=8254269(8.3M)  fh1=94%  pred=0  nps=8.2M
        chk=485.7K  qchk=493.6K  fp=3.7M  mcp=15.6K  50move=1  egtb=103.8K
        LMReductions:  1/430.5K  2/247.8K  3/87.9K  4/14.2K  5/141
        null-move (R):  3/17.6K  4/11.4K  5/2.3K  6/603  7/38
White(1): Kb1
              time used:   1.01
Black(1):
another one:

White(1): [d]8/1p3pp1/7p/5P1P/2k3P1/8/2K2P2/8 w - - 0 1

Code: Select all

White(1): g
        time surplus   0.00  time limit 2.00 (2.00)
        depth     time       score   variation (1)
          ....
          7     0.00/2.00    -1.25   1. f3 b5 2. Kb2 b4 3. Kc2 f6 4. Kb2
          7->   0.00/2.00    -1.25   1. f3 b5 2. Kb2 b4 3. Kc2 f6 4. Kb2
          8     0.00/2.00    -1.27   1. f3 b5 2. Kb2 b4 3. Kc2 b3+ 4. Kb2 f6
                                     5. f4
          8->   0.00/2.00    -1.27   1. f3 b5 2. Kb2 b4 3. Kc2 b3+ 4. Kb2 f6
                                     5. f4
          9     0.00/2.00    -1.36   1. f3 Kd4 2. g5 f6 3. gxf6 gxf6 4. Kb3 Ke3
                                     5. Kc4 Kxf3
          9     0.00/2.00    -1.11   1. Kd2 b5 2. Kc2 b4 3. f3 Kd4 4. Kb2 Ke3
                                     5. Kb3 Kxf3 6. Kxb4 Kxg4
          9->   0.00/2.00    -1.11   1. Kd2 b5 2. Kc2 b4 3. f3 Kd4 4. Kb2 Ke3
                                     5. Kb3 Kxf3 6. Kxb4 Kxg4
         10     0.00/2.00    -1.11   1. Kd2 b5 2. Kc2 b4 3. f3 Kd4 4. Kb2 Ke3
                                     5. Kb3 Kxf3 6. Kxb4 Kxg4
         10->   0.00/2.00    -1.11   1. Kd2 b5 2. Kc2 b4 3. f3 Kd4 4. Kb2 Ke3
                                     5. Kb3 Kxf3 6. Kxb4 Kxg4
         11     0.01/2.00    -1.06   1. Kd2 b5 2. Kc2 b4 3. g5 f6 4. g6 Kd4
                                     5. Kb3 Kc5 6. f3
         11->   0.01/2.00    -1.06   1. Kd2 b5 2. Kc2 b4 3. g5 f6 4. g6 Kd4
                                     5. Kb3 Kc5 6. f3
         12     0.01/2.00    -1.11   1. Kd2 b5 2. Kc2 b4 3. g5 f6 4. gxf6 b3+
                                     5. Kb2 gxf6 6. f3 Kb4 7. f4
         12     0.02/2.00     5.92   1. f6 gxf6 2. f4 Kd4 3. g5 hxg5 4. fxg5 f5
                                     5. h6 f4 6. h7 f3 7. h8=Q+ Ke3
         12->   0.02/2.00     5.92   1. f6 gxf6 2. f4 Kd4 3. g5 hxg5 4. fxg5 f5
                                     5. h6 f4 6. h7 f3 7. h8=Q+ Ke3
         13     0.02/2.00     6.66   1. f6 gxf6 2. f4 Kd5 3. g5 hxg5 4. fxg5
                                     Ke6 5. h6 f5 6. h7 f6 7. h8=Q fxg5
         13->   0.02/2.00     6.66   1. f6 gxf6 2. f4 Kd5 3. g5 hxg5 4. fxg5
                                     Ke6 5. h6 f5 6. h7 f6 7. h8=Q fxg5
         ...         
         18     0.37/2.00    16.71   1. f6 gxf6 2. f4 Kd5 3. g5 hxg5 4. fxg5
                                     Ke6 5. h6 f5 6. g6 f4 7. h7 f6 8. g7 Kf5
                                     9. g8=Q f3 10. h8=Q
         18->   0.38/2.00    16.71   1. f6 gxf6 2. f4 Kd5 3. g5 hxg5 4. fxg5
                                     Ke6 5. h6 f5 6. g6 f4 7. h7 f6 8. g7 Kf5
                                     9. g8=Q f3 10. h8=Q
         19     1.36/2.00     Won    1. f6 gxf6 2. f4 Kd5 3. g5 hxg5 4. fxg5
                                     Ke6 5. h6 f5 6. h7 Kd7 7. h8=Q b6 8. Qf8
                                     Ke6 9. Qe8+ Kd6 10. Qxf7 <EGTB>
         19->   1.36/2.00     Won    1. f6 gxf6 2. f4 Kd5 3. g5 hxg5 4. fxg5
                                     Ke6 5. h6 f5 6. h7 Kd7 7. h8=Q b6 8. Qf8
                                     Ke6 9. Qe8+ Kd6 10. Qxf7 <EGTB>
         ...
         24     1.94/2.00     Won    1. f6 gxf6 2. f4 Kd5 3. g5 hxg5 4. fxg5
                                     Ke6 5. h6 f5 6. h7 Kd7 7. h8=Q b6 8. Qf8
                                     Ke6 9. Qe8+ Kd6 10. Qxf7 <EGTB>
         24->   1.98/2.00     Won    1. f6 gxf6 2. f4 Kd5 3. g5 hxg5 4. fxg5
                                     Ke6 5. h6 f5 6. h7 Kd7 7. h8=Q b6 8. Qf8
                                     Ke6 9. Qe8+ Kd6 10. Qxf7 <EGTB>
        time=2.00(100%)  nodes=15704290(15.7M)  fh1=90%  pred=0  nps=7.9M
        chk=932.9K  qchk=1.7M  fp=5.2M  mcp=94.6K  50move=0  egtb=110.4K
        LMReductions:  1/290.0K  2/147.4K  3/34.7K  4/3.3K  5/4
        null-move (R):  3/52.6K  4/25.9K  5/683  6/7
White(1): f6
              time used:   2.00
Black(1):
Top
User avatar
MikeB
Posts: 4889
Joined: Thu Mar 09, 2006 6:34 am
Location: Pen Argyl, Pennsylvania

Re: Latest Crafty-25.1-ACCA

Post by MikeB »

here are the bench score with 256M hash and 64M hashp (pawn hash)

Code: Select all

White(1): bench
Running serial benchmark. . .
pos:  1(7.4M)  2(6.2M)  3(6.9M)  4(7.3M)  5(5.7M)  6(6.6M)  7(988.3K) 
pos:  8(6.0M)  9(6.7M) 10(6.6M) 11(6.1M) 12(6.8M) 13(5.7M) 14(6.5M) 
pos: 15(11.9M) 16(6.7M) 17(4.5M) 18(5.7M) 19(5.8M) 20(7.1M) 21(8.4M) 
pos: 22(6.6M) 23(6.2M) 24(2.2M) 25(7.0M) 26(5.7M) 27(2.7K) 28(5.7M) 
pos: 29(4.7M) 30(6.9M) 31(6.1M) 32(6.0M) 33(6.3M) 34(6.2M) 35(4.3M) 
pos: 36(6.7M) 37(6.7M) 38(1.6K) 39(6.4M) 40(2.4K) 41(6.6M) 42(7.2M) 
pos: 43(6.2M) 44(7.2M) 45(6.3M) 46(6.6M) 47(6.3M) 48(6.5M) 49(9.0M) 
pos: 50(6.2M) 51(7.5M) 52(6.7M) 53(5.3M) 54(1.6M) 55(7.2M) 56(5.2M) 
pos: 57(8.4M) 58(6.9M) 59(939.3K) 60(7.0M) 61(7.2M) 62(1.1K) 63(6.3M) 
pos: 64(6.2M) 

Total nodes: 176051914
Raw nodes per second: 6192469
Total elapsed time: 28.43

Running SMP benchmark (2 threads)...
pos:  1(13.1M)  2(11.3M) 

Total nodes: 179242160
Raw nodes per second: 6245371
Total elapsed time: 28.70
here is the "bench 6" increasing depth by 6 plies, with 12 cores

Code: Select all

White(1): mt 12
max threads set to 12.
White(1): bench 6
Running serial benchmark (modifying depth by 6 plies). . .
pos:  1(37.9M)  2(39.3M)  3(48.6M)  4(43.6M)  5(42.5M)  6(54.8M)  7(4.9M) 
pos:  8(31.5M)  9(51.6M) 10(30.4M) 11(33.4M) 12(56.4M) 13(13.8M) 14(52.5M) 
pos: 15(105.8M) 16(44.0M) 17(26.1M) 18(50.0M) 19(52.8M) 20(13.9M) 21(53.6M) 
pos: 22(46.5M) 23(27.1M) 24(25.6M) 25(59.3M) 26(6.2M) 27(2.7K) 28(44.9M) 
pos: 29(25.9M) 30(40.4M) 31(50.4M) 32(55.4M) 33(52.9M) 34(26.3M) 35(22.0M) 
pos: 36(61.2M) 37(53.0M) 38(1.6K) 39(57.6M) 40(2.4K) 41(44.0M) 42(61.3M) 
pos: 43(53.0M) 44(53.1M) 45(42.8M) 46(61.6M) 47(53.6M) 48(32.5M) 49(12.2M) 
pos: 50(52.0M) 51(52.8M) 52(43.0M) 53(41.5M) 54(3.2M) 55(57.4M) 56(34.8M) 
pos: 57(23.0M) 58(48.4M) 59(4.3M) 60(44.2M) 61(44.6M) 62(1.1K) 63(59.6M) 
pos: 64(50.3M) 

Total nodes: 3749301043
Raw nodes per second: 51721631
Total elapsed time: 72.49

Running SMP benchmark (2 threads)...
pos: 

Total nodes: 3749301043
Raw nodes per second: 51721631
Total elapsed time: 72.49
here is bench 12 , increasing depth by 12 plies with 12 cores - here you see the added benefit of 12 cores on a deeper search , averaging just about 62,000,000 nps.

Code: Select all

White(1): bench 12
Running serial benchmark (modifying depth by 12 plies). . .
pos:  1(49.8M)  2(56.5M)  3(63.3M)  4(57.2M)  5(51.7M)  6(63.7M)  7(10.7M) 
pos:  8(42.7M)  9(57.0M) 10(44.0M) 11(40.1M) 12(59.0M) 13(29.9M) 14(62.7M) 
pos: 15(99.5M) 16(51.7M) 17(30.8M) 18(57.6M) 19(59.5M) 20(21.8M) 21(55.1M) 
pos: 22(58.6M) 23(30.6M) 24(34.0M) 25(67.8M) 26(6.1M) 27(2.7K) 28(51.7M) 
pos: 29(32.0M) 30(59.2M) 31(60.9M) 32(61.6M) 33(63.9M) 34(43.0M) 35(34.0M) 
pos: 36(68.6M) 37(65.6M) 38(1.6K) 39(63.9M) 40(2.4K) 41(61.1M) 42(70.7M) 
pos: 43(56.1M) 44(69.9M) 45(45.5M) 46(65.1M) 47(63.3M) 48(40.3M) 49(13.8M) 
pos: 50(62.4M) 51(68.0M) 52(59.0M) 53(57.7M) 54(4.4M) 55(61.7M) 56(35.7M) 
pos: 57(23.5M) 58(61.1M) 59(14.8M) 60(56.7M) 61(55.4M) 62(1.1K) 63(66.3M) 
pos: 64(61.9M) 

Total nodes: 49279251806
Raw nodes per second: 61969331
Total elapsed time: 795.22

Running SMP benchmark (2 threads)...
pos: 

Total nodes: 49279251806
Raw nodes per second: 61969331
Total elapsed time: 795.22

White(1):
Top
jdart
Posts: 4435
Joined: Fri Mar 10, 2006 5:23 am
Location: http://www.arasanchess.org

Re: Latest Crafty-25.1-ACCA

Post by jdart »

I saw Crafty hitting over 90m nps at one point in the tournament.

--Jon
Top
CRoberson
Posts: 2096
Joined: Mon Mar 13, 2006 2:31 am
Location: North Carolina, USA

Re: Latest Crafty-25.1-ACCA

Post by CRoberson »

Crafty is playing Telepath now and Crafty's peak NPS is 111.2 M NPS !
The lowest I saw so far is 88.6 M NPS.
Top
Rochester
Posts: 55
Joined: Sat Feb 20, 2016 6:11 am

Re: Latest Crafty-25.1-ACCA

Post by Rochester »

The news is too good!! crafty can be the best yes? Long time no see happen, thank you!
Top
carldaman
Posts: 2287
Joined: Sat Jun 02, 2012 2:13 am

Re: Latest Crafty-25.1-ACCA

Post by carldaman »

jdart wrote:I saw Crafty hitting over 90m nps at one point in the tournament.

--Jon
You can get a speeding ticket for that. :P
Top
bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: Latest Crafty-25.1-ACCA

Post by bob »

jdart wrote:I saw Crafty hitting over 90m nps at one point in the tournament.

--Jon
Note that Mike is running on his 12 core Mac. I am using the same 20 core ES2660 @ 2.9ghz I have been using for the last 2-3 years. It seems to hit between 80M and 100M. More for longer games.
Top
Paloma
Posts: 1235
Joined: Thu Dec 25, 2008 9:07 pm
Full name: Herbert L

Re: Latest Crafty-25.1-ACCA

Post by Paloma »

Is Crafty-25.1 for Windows available ?
Top
bob
Posts: 20943
Joined: Mon Feb 27, 2006 7:30 pm
Location: Birmingham, AL

Re: Latest Crafty-25.1-ACCA

Post by bob »

Paloma wrote:Is Crafty-25.1 for Windows available ?
Won't be long.
Top
Joerg Oster
Posts: 996
Joined: Fri Mar 10, 2006 4:29 pm
Location: Germany
Full name: Jörg Oster

Re: Latest Crafty-25.1-ACCA

Post by Joerg Oster »

bob wrote:
Paloma wrote:Is Crafty-25.1 for Windows available ?
Won't be long.
Is the source code available somewhere?
I'd love to take a peek at your SMP implementation. :D
Jörg Oster
Top

Return to “Computer Chess Club: General Topics”