MANIAC I

From Chessprogramming wiki
Jump to: navigation, search

Home * Engines * MANIAC I

MANIAC I,
the chess program on a MANIAC I (Mathematical Analyzer, Numerical Integrator, and Computer or Mathematical Analyzer, Numerator, Integrator, and Computer), the machine designed and build by a team around John von Neumann and Nicholas Metropolis at the Los Alamos Scientific Laboratory. The MANIAC I chess program was written in 1956 by a group of H-bomb researchers, Stanislaw Ulam, Paul Stein, Mark Wells, James Kister, William Walden and John Pasta. Due to lack of computing power, only a chess variant with a reduced 6 x 6 board was implemented, without bishops, double-step for pawns and castling , later called Los Alamos Chess.

Photos

2-2.MANIAC.LAT1994.L02645387.LANL.lg.jpg

Los Alamos scientists Paul Stein (left) and Nick Metropolis playing chess with the MANIAC computer [1]

Description

MANIAC I performed a brute-force Shannon Type A strategy, pure minimax. During game play with 11,000 ops./sec, it searched 4 plies deep in about 12 minutes to find its best move. The program was written in 600 words of machine code. Its evaluation took material and mobility under account, both incrementally updated during make and unmake move [2].

Quotes

Quote from Chronology of Computing compiled by David Singmaster [3]

A group at Los Alamos, based on Kister, Stein, Ulam, Walden and Wells, follows up a brief Russian reference to a chess program for BESM [4]. The Los Alamos group writes a program for the MANIAC I to play a reduced game of chess – using a 6 x 6 board without bishops. 

Fred Guterl

Quote by Fred Guterl from Silicon gambit [5] :

The government laboratory in Los Alamos, New Mexico, got hold of one of the first computers, MANIAC I, so that Ulam and the other H-bomb researchers wouldn't have to stay up nights solving their voluminous equations with pencil and paper. Ulam, who described himself modestly as a "fair" chess player, couldn't resist putting the machine to work on a project of somewhat less import to coldwar strategy. Together with physicist Paul Stein, he wrote one of the first chess-playing programs. 

Roger Snodgrass

Roger Snodgrass in LANL: The Rest of the Story on MANIAC and Mark Wells [6]

Among the interesting tidbits in Wells article are stories about a chess-playing program on MANIAC. MANIAC’s limited memory restricted a play to board that was six squares by six squares and no bishops...
“Even then,” he wrote, “moves averaged about 10 minutes for a two-move, look-ahead strategy.” “That quickly became three moves, four moves, five moves ahead,” Wells said Tuesday, adding the current capability was at least 12 moves ahead. 
His essay also includes an anecdote about a moment when the computer seemed to have a mind of its own. When Princeton physicist Martin Kruskal checkmated the MANIAC on the 38th move of a game, the machine responded with an illegal move. “We were dumbfounded for a while, until we traced the trouble and realized that the program had never been taught to resign,” Wells wrote. Facing no moves, the machine was stuck in a loop and the loop changed the program.
“You might call that a learning program,” he recalled. 

Selected Games

MANIAC I played a game against a young lady who had learnt the game a week earlier. It was the first time a human had lost to a computer in a game of intellectual skill [7]:

[Event "6x6 Los Alamos Chess"]
[Site "Los Alamos"]
[Date "1956.??.??"]
[Round "?"]
[White "MANIAC I"]
[Black "Human"]
[Result "1-0"]

1.d3 b4 2.Nf3 d4 3.b3 e4 4.Ne1 a4 5.bxa4 Nxa4 6.Kd2 Nc3 7.Nxc3 bxc3+ 8.Kd1 f4 
9.a3 Rb6 10.a4 Ra6 11.a5 Kd5 12.Qa3 Qb5 13.Qa2+ Ke5 14.Rb1 Rxa5 15.Rxb5 Rxa2 
16.Rb1 Ra5 17.f3 Ra4 18.fxe4 c4 19.Nf3+ Kd6 20.e5+ Kd5 21.exf6Q Nc5 22.Qf6xd4+ 
Kc6 23.Nf3-e5 1-0

See also

Selected Publications

External Links

Chess Program

Computer

Misc

Wolfgang Schlüter, Volker Kriegel, Curt Cress, Hans Peter Ströer, Nippy Noya, Alan Skidmore, Rainer Brüninghaus

References

  1. Los Alamos scientisits Paul Stern (left) and Nick Metropolis playing chess with the MANIAC computer, 1956, Courtesy of Los Alamos National Laboratory, hosted by The Computer History Museum
  2. Allen Newell, Cliff Shaw, Herbert Simon (1958). Chess Playing Programs and the Problem of Complexity. IBM Journal of Research and Development, Vol. 4, No. 2, pp. 320-335. Reprinted (1963) in Computers and Thought (eds. Edward Feigenbaum and Julian Feldman), pp. 39-70. McGraw-Hill, pdf, pp. 45 Table I Comparison of Current Chess Programs
  3. Chronology of Computing compiled by David Singmaster
  4. "There are two other explorations between 1951 and 1956 of which we are aware - a hand simulation by F. Mosteller and a Russian program for BESM. Unfortunately, not enough information is available on either to talk about them, so we must leave a gap in the history between 1951 and 1956" - footnote 1 in Allen Newell, Cliff Shaw, Herbert Simon (1958). Chess Playing Programs and the Problem of Complexity. IBM Journal of Research and Development, Vol. 4, No. 2, Reprinted (1963) in Computers and Thought (eds. Edward Feigenbaum and Julian Feldman), pp. 47. McGraw-Hill, pdf
  5. Silicon gambit by Fred Guterl, Discover, June 01, 1996
  6. LANL: The Rest of the Story by Roger Snodgrass, Los Alamos Monitor Editor, July 16, 2008
  7. From the Z1 to the Singularity – Zuse's 100th birthday by Frederic Friedel, ChessBase News, June 22, 2010
  8. Re: Old programs CHAOS and USC by Dann Corbit, CCC, July 11, 2015
  9. from Martin H. Weik (1961). A Third Survey of Domestic Electronic Digital Computing Systems. Report No. 1115

Up one level