In a repeat of man versus machine, world chess champion Garry Kasparov will today square off once again against IBM's Deep Blue supercomputer.
Kasparov won the first six-game match in February 1996, 4-2.
Although he will be facing a faster and improved version of IBM's Deep Blue RS/6000 chess computer, Kasparov is confident that he will win again.
"It should be more difficult this time, but computers are always beatable," Kasparov told a news conference in New York. "It's billions of calculations against my intuition."
Conceding that Kasparov is at the height of his career, IBM's Deep Blue team this year is going against Kasparov with a faster and more intelligent version of the supercomputer.
"The processor is twice as fast ... and it can analyze 200 million chess positions per second," says Feng Hsiung Hsu, lead scientist with the Deep Blue at IBM's Thomas J. Watson Research Center in New York.
Kasparov can examine approximately three positions per second, IBM officials say.
The latest iteration of Deep Blue is a 32-node IBM RS/6000 SP high-performance computer, built around the new Power Two Super Chip processors (P2SC). Each node of the SP employs a single microchannel card containing eight dedicated VLSI chess processors, for a total of 256 processors working in tandem, IBM officials said.
Deep Blue's programming code is written in C and runs under the AIX operating system.
The supercomputer has gained a lot of chess strategy since it last went up against its human rival 15 months ago. For instance, Kasparov was able to win last year partly because he shifted strategy in midgame, confusing Deep Blue in the process.
"The outcome of the match partly depends on Deep Blue's ability to shift strategy and on my ability to concentrate," Kasparov said.
The six-game match will begin Saturday at the Equitable Center at 7th Avenue and 51st Street. The winner will take home US$700,000 of a $1.1 million purse.
The game can be followed on the Internet at http://www.chess.ibm.com/. The IBM home page can be found on the Internet at http://www.ibm.com/. IBM's Research home page can be found at http://www.research.ibm.com/.