Any Examples of The King's Tactical Abilities?

[Werewolf]

Hi Steve,
in a chess computers competition I recently ran, I have been impressed by The King Performance announcement 'mate in 13', level Normal 5 (average 15secs/move) after 5 seconds and 270 Knodes:

[d]6k1/qp4pb/2p4p/p1n5/2P1N3/P2rQP1P/1P4P1/1K2R3 w - - 0 29

The engine is The King v2.61; CPU: ARM Cortex M7 @300Mhz.

I checked using several modern engines on my PC (Intel Core i5), many are way slower.
As you suspect, The King is very efficient at mating tactics.
Warm regards,
Tibono

What nps do you get from the start position on that machine?
I think the engine is basically the same as the TASC R30 1995

[Tibono]

What nps do you get from the start position on that machine?

Close to 30 Knodes/sec. Displayed steps (nodes & time):

Code: Select all

919826   31secs --> 29671
3323K  1m51secs --> 29937
4421K  2m29secs --> 29671
11737K 6m26secs --> 30407

KR,
Tibono

[Rebel]

[pgn][Event ""]
[Site ""]
[Date "1994.1.26"]
[White "The King"]
[Black ""]
[Result "1/2-1/2"]
[Setup "1"]
[FEN "r7/8/4p2k/3p1bR1/7r/Np6/1P6/K5R1 w - - 0 1"]

1. Rh5+ *[/pgn]
r7/8/4p2k/3p1bR1/7r/Np6/1P6/K5R1 w - -

If memory serves me well this was during the WCCC in Munich 1993 and was unprecedented with the hardware of that time.

[Tibono]

If memory serves me well this was during the WCCC in Munich 1993 and was unprecedented with the hardware of that time.

Thank you Ed for this test position; well, 2019 King Performance plays the same Rh5+ move and announces 0.00 score in 10 secs and 392Knodes.
Nice!

[JVMerlino]

[d]8/6B1/3p4/3P2pk/5q1p/N2Q3b/3n2P1/6K1 b - - 0 1

Mephisto Almeria - Deep Thought
The supercomputer found 1...Bg2! at tournament time controls. The TASC R40 finds it in 20 minutes.

The King from CM9000 finds Bxg2 in less than second.

[JVMerlino]

From Capablanca's notes to one of his games:

1. e4 e6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Nd7 5. Nf3 Ngf6 6. Nxf6+ Nxf6 7. Ne5 Bd6 8. Qf3 c6 9. c3 O-O 10. Bg5 Be7 11. Bd3 Bd7 12. Qh3 h6

[d]r2q1rk1/pp1bbpp1/2p1pn1p/4N1B1/3P4/2PB3Q/PP3PPP/R3K2R w KQ - 0 13
13. Bxh6 !

The King from CM9000 finds Bxh6 in one second.

[JVMerlino]

And one more:

1. e4 Nf6 2. e5 Nd5 3. c4 Nb6 4. d4 Nc6

[d]r1bqkb1r/pppppppp/1nn5/4P3/2PP4/8/PP3PPP/RNBQKBNR w KQkq - 0 5

5. d5!

The King from CM9000 finds this "instantly", needing to search only 57,656 nodes at depth 1/6.

[JVMerlino]

Hi Steve,
in a chess computers competition I recently ran, I have been impressed by The King Performance announcement 'mate in 13', level Normal 5 (average 15secs/move) after 5 seconds and 270 Knodes:

[d]6k1/qp4pb/2p4p/p1n5/2P1N3/P2rQP1P/1P4P1/1K2R3 w - - 0 29

The engine is The King v2.61; CPU: ARM Cortex M7 @300Mhz.

The King from CM9000 (v3.23) announces Mate in 13 in less than a second.

[Werewolf]

[pgn][Event ""]
[Site ""]
[Date "1994.1.26"]
[White "The King"]
[Black ""]
[Result "1/2-1/2"]
[Setup "1"]
[FEN "r7/8/4p2k/3p1bR1/7r/Np6/1P6/K5R1 w - - 0 1"]

1. Rh5+ *[/pgn]
r7/8/4p2k/3p1bR1/7r/Np6/1P6/K5R1 w - -

If memory serves me well this was during the WCCC in Munich 1993 and was unprecedented with the hardware of that time.

One thing I never understood is why Mephisto didn't realise a 32MHz version of Gideon 3.1 (RISC II program - which ran at 14MHz) as TASC did with The ChessMachine.

[Rebel]

[pgn][Event ""]
[Site ""]
[Date "1994.1.26"]
[White "The King"]
[Black ""]
[Result "1/2-1/2"]
[Setup "1"]
[FEN "r7/8/4p2k/3p1bR1/7r/Np6/1P6/K5R1 w - - 0 1"]

1. Rh5+ *[/pgn]
r7/8/4p2k/3p1bR1/7r/Np6/1P6/K5R1 w - -

If memory serves me well this was during the WCCC in Munich 1993 and was unprecedented with the hardware of that time.

One thing I never understood is why Mephisto didn't realise a 32MHz version of Gideon 3.1 (RISC II program - which ran at 14MHz) as TASC did with The ChessMachine.

They were already close to bankruptcy, so it did not matter any longer.