CHESS

The author of the Rybka program, Vasik Rajlich, made some startling claims with regards to the soundness of the Kings Gambit (1 e4 e5 2 f4 exf4). He had harnessed an immense amount of computing power and focused it on this most romantic of chess openings, and after a number of months has reached the following conclusion: ‘We now know the exact outcome of this position, assuming perfect play, of course. Curiously there is only one move that draws for White, and that is, somewhat surprisingly, 3.Be2. Every other move loses by force. We set up a cluster of computers, currently around 300 cores, which has been used by World Champions and World Champion candidates to prepare for their matches. It is arguably the most powerful entity to play chess, ever, anywhere. Well, that was until we hooked it up to a massively parallel cluster of IBM POWER 7 Servers provided by IBM’s Semantic Analysis and Integration department – 2,880 cores at 4.25 GHz, 16 terabytes of RAM. We developed an algorithm, which attempts to classify chess positions into wins, draws and losses. Using this algorithm, we have just finished classifying the King’s Gambit. In other words, the King’s Gambit is now solved.’ While the number of moves Rybka looks at is truly astronomical, it at the same time makes some prudent shortcuts by not actually playing every conceivable line down to mate. Once an evaluation reaches +/- 5.12 it assumes there is a mate further down the tree and continues its search down another line. Over 50 years ago Bobby Fischer published a famous article claiming to have refuted the Kings Gambit, and the basis of it has been backed up by Rybka’a analysis

and Ficher’s choice on moves 3-5 have all got the silicon thumbs up!’ Note: The above is an abbreviated version of an article on the chessbase website that was released suspiciously close to April 1…