Stockfish 1.7.1 64-bit 4CPU - gauntlet (CCRL 40/40)

[Martin Thoresen]

Stockfish 1.7.1 64-bit 4CPU Gauntlet

CPU: Intel i7 920 @ 4101 MHz
GUI: Chessbase 12
Hash: 1024 MB
Nalimov Tablebases: 3-4-5
Ponder: Off
Openings: Modern SGM v3.ctg (Limited To 8 Moves)
Time Control: 40 Moves In 17 Minutes Repeating (Equivalent To CCRL 40/40)
Format: Gauntlet
Length: 50 Games Per Opponent
Total Number Of Games: 400

Code: Select all

Stockfish_171_Gauntlet_CCRL  2010

Stockfish 1.7.1 64-bit 4CPU   - Deep Fritz 12 4CPU
Stockfish 1.7.1 64-bit 4CPU   - Deep Shredder 12 64-bit 4CPU
Stockfish 1.7.1 64-bit 4CPU   - Deep Sjeng WC2008 64-bit 4CPU
Stockfish 1.7.1 64-bit 4CPU   - Komodo 1.0 64-bit
Stockfish 1.7.1 64-bit 4CPU   - Naum 4.2 64-bit 4CPU
Stockfish 1.7.1 64-bit 4CPU   - Rybka 3 64-bit 4CPU
Stockfish 1.7.1 64-bit 4CPU   - Spark 0.4 64-bit 4CPU
Stockfish 1.7.1 64-bit 4CPU   - Zappa Mexico II 64-bit 4CPU

[El Gringo]

Hi Martin,


How do you keep the temperatures under control with all the heat that the i7 920 produces at 4.1GHz.

Watercooling ,aircooling ?

Best
Johan

[Martin Thoresen]

Hi Martin,


How do you keep the temperatures under control with all the heat that the i7 920 produces at 4.1GHz.

Watercooling ,aircooling ?

Best
Johan

Hello Johan,

This is a fully air cooled computer. Peak temperature during 100% load is
74 degrees Centigrade with Core Temp, so it's at the limit of what I think is
acceptable.

[Martin Thoresen]

Standings after 56 of 400 games:

Code: Select all

Stockfish_171_Gauntlet_CCRL  2010

Stockfish 1.7.1 64-bit 4CPU   - Deep Fritz 12                     5.5 - 1.5    +4/=3/-0    78.57%
Stockfish 1.7.1 64-bit 4CPU   - Deep Shredder 12 64-bit 4CPU      5.5 - 1.5    +4/=3/-0    78.57%
Stockfish 1.7.1 64-bit 4CPU   - Deep Sjeng WC2008 64-bit 4CPU     6.5 - 0.5    +6/=1/-0    92.86%
Stockfish 1.7.1 64-bit 4CPU   - Komodo 1.0 64-bit                 5.5 - 1.5    +4/=3/-0    78.57%
Stockfish 1.7.1 64-bit 4CPU   - Naum 4.2 64-bit 4CPU              3.5 - 3.5    +2/=3/-2    50.00%
Stockfish 1.7.1 64-bit 4CPU   - Rybka 3 64-bit 4CPU               4.0 - 3.0    +2/=4/-1    57.14%
Stockfish 1.7.1 64-bit 4CPU   - Spark 0.4 64-bit 4CPU             6.0 - 1.0    +5/=2/-0    85.71%
Stockfish 1.7.1 64-bit 4CPU   - Zappa Mexico II 64-bit 4CPU       5.5 - 1.5    +4/=3/-0    78.57%

Download games: 1-56

[ArmyBridge]

Very very impressed after 56 games , face all this monster (Rybka 3 4 CPU included) and NO SINGLE LOSE fantastic engine

Amazing Regards

[Martin Thoresen]

After 80 of 400 games:

Code: Select all

Stockfish_171_Gauntlet_CCRL  2010

Stockfish 1.7.1 64-bit 4CPU   - Deep Fritz 12 4CPU                8.0 - 2.0    +6/=4/-0    80.00%
Stockfish 1.7.1 64-bit 4CPU   - Deep Shredder 12 64-bit 4CPU      7.5 - 2.5    +5/=5/-0    75.00%
Stockfish 1.7.1 64-bit 4CPU   - Deep Sjeng WC2008 64-bit 4CPU     8.0 - 2.0    +7/=2/-1    80.00%
Stockfish 1.7.1 64-bit 4CPU   - Komodo 1.0 64-bit                 8.5 - 1.5    +7/=3/-0    85.00%
Stockfish 1.7.1 64-bit 4CPU   - Naum 4.2 64-bit 4CPU              6.0 - 4.0    +4/=4/-2    60.00%
Stockfish 1.7.1 64-bit 4CPU   - Rybka 3 64-bit 4CPU               5.5 - 4.5    +2/=7/-1    55.00%
Stockfish 1.7.1 64-bit 4CPU   - Spark 0.4 64-bit 4CPU             8.5 - 1.5    +7/=3/-0    85.00%
Stockfish 1.7.1 64-bit 4CPU   - Zappa Mexico II 64-bit 4CPU       7.5 - 2.5    +5/=5/-0    75.00%

Games will be uploaded at a later time.

[Tord Romstad]

This was a fun game: You don't often see top programs getting mated in the middle of the board in less than 20 moves these days, particularly not when playing white at a slow time control:

Code: Select all

[Event "Stockfish_171_Gauntlet_CCRL"]
[Site "Norway"]
[Date "2010.04.12"]
[Round "2.1"]
[White "Deep Fritz 12 4CPU"]
[Black "Stockfish 1.7.1 64-bit 4CPU"]
[Result "0-1"]
[ECO "C45"]
[Annotator "0.07;-1.01"]
[PlyCount "30"]
[EventDate "2010.04.12"]
[EventType "simul"]
[Source "Thoresen"]
[TimeControl "40/1020:0/0:0/0"]

{Intel(R) Core(TM) i7 CPU         920  @ 2.67GHz 3906 MHz  W=17.3 plies; 15
 556kN/s; Modern SGM v3.ctg  B=23.3 plies; 8 272kN/s; Modern SGM v3.ctg} 1. e4
{B 0} e5 {B 0} 2. Nf3 {B 0} Nc6 {B 0} 3. d4 {B 0} exd4 {B 0} 4. Nxd4 {B 0} Bc5
{B 0} 5. Nb3 {B 0} Bb6 {B 0} 6. Nc3 {B 0} Nf6 {B 0} 7. Bg5 {B 0} h6 {B 0} 8.
Bh4 {B 0} d6 {B 0} 9. a4 {0.07/20 29} Nxe4 {-1.01/24 37 (a5)} 10. Bxd8 {-1.30/
19 20 (Qh5)} Bxf2+ {-9.81/24 47} 11. Ke2 {-0.23/13 0} Bg4+ {-9.85/21 3} 12. Kd3
{0.04/13 0} Ne5+ {-10.22/24 39} 13. Kxe4 {-1.18/13 0} f5+ {-10.90/23 37} 14.
Kd5 {-6.74/22 32} Rxd8 {-10.50/23 34} 15. Nd4 {-8.12/21 30 (Qxg4)} c6+ {-10.90/
24 39} 0-1

Stockfish sacrifices the queen with 9... Nxe4!:
[d]r1bqk2r/ppp2pp1/1bnp1n1p/8/P3P2B/1NN5/1PP2PPP/R2QKB1R b KQkq - 0 9

Final position, after 15... c6+:
[d]3rk2r/pp4p1/2pp3p/3Knp2/P2N2b1/2N5/1PP2bPP/R2Q1B1R w - 0 16

[alpha123]

Stockfish is starting to play like Fritz.... (DF10.1 anyway....)

Shows how tactically weak and boring the newer Fritzes are.

Congratulations on Stockfish's performance by the way.

Peter

[Martin Thoresen]

Tord,

I am not sure that game is very valid, seems to be a dubious
opening if you ask me. But fun, oh yeah! It might get filtrated
out of the CCRL database when I upload it for the rating list,
I am not sure however since I think I can remember that both
engines have to have a evaluation out of book of less than one
half pawn; in this case DF 12 is almost at zero but on the other
hand Stockfish is claiming over a pawn advantage out of book.

Download games 57-80

[Graham Banks]

Tord,

I am not sure that game is very valid, seems to be a dubious
opening if you ask me. But fun, oh yeah! It might get filtrated
out of the CCRL database when I upload it for the rating list,
I am not sure however since I think I can remember that both
engines have to have a evaluation out of book of less than one
half pawn; in this case DF 12 is almost at zero but on the other
hand Stockfish is claiming over a pawn advantage out of book.

Download games 57-80

I think that the game will be okay. It's only if both sides believe that the winning engine has an advantage of 0.70 or more for the entire game that it's considered a bad opening.

Cheers,
Graham.