Deep Blue vs Kasparov (1997) was a six-game chess match played in New York City from May 3 to May 11, 1997, between world champion Garry Kasparov and IBM's chess computer Deep Blue. The match ended with Deep Blue winning 3.5 to 2.5, making it the first time a computer defeated a reigning world chess champion in a match under standard tournament time controls. The event was widely covered by international media and is considered a landmark moment in the history of Artificial intelligence and computing.
The match was a rematch of a 1996 contest in which Kasparov had defeated Deep Blue 4-2. For the 1997 match, IBM had significantly upgraded Deep Blue's hardware and software, doubling its processing power and improving its chess evaluation functions. Deep Blue was a specialized Chess computer capable of evaluating up to 200 million positions per second, using custom very-large-scale integration (VLSI) chips and a parallel architecture.
Match Summary
The match consisted of six games, with the first player to score 3.5 points declared the winner. Kasparov won the first game, but Deep Blue won the second, marking the first time a computer had ever beaten Kasparov in a tournament game. Games three, four, and five were all draws. In the final game, played on May 11, Kasparov made a critical error early and resigned after only 19 moves, giving Deep Blue the match victory.
Game-by-game results: Game 1 (May 3): Kasparov (White) defeated Deep Blue. Game 2 (May 4): Deep Blue (White) defeated Kasparov. Game 3 (May 6): Draw. Game 4 (May 7): Draw. Game 5 (May 9): Draw. Game 6 (May 11): Deep Blue (White) defeated Kasparov.
Technical Background
Deep Blue was developed by a team at IBM led by Feng-hsiung Hsu, with contributions from Murray Campbell, A. Joseph Hoane, and others. The system combined brute-force search with sophisticated evaluation functions and opening book knowledge. Unlike modern Machine learning approaches, Deep Blue did not use Neural network or Deep learning techniques; it relied on hand-crafted evaluation heuristics and massive parallel search.
The computer's hardware consisted of 30 IBM RS/6000 SP nodes, each containing 16 custom chess chips, for a total of 480 processors working in parallel. This architecture allowed Deep Blue to search to depths of 10 to 12 plies (half-moves) in normal positions, and much deeper in tactical situations. The system also had access to a database of hundreds of thousands of grandmaster games for its opening book.
Aftermath and Significance
The victory was controversial. Kasparov alleged that IBM had cheated by having human grandmasters intervene during the games, a claim IBM denied. IBM refused a rematch and retired Deep Blue shortly after the match, donating one of its racks to the Smithsonian Institution. The match sparked widespread debate about the capabilities of computers relative to human intelligence and influenced public perception of Artificial intelligence.
In the years following, the event is often cited as a milestone in computing history, though it is distinct from later advances in Machine learning and Deep learning. Deep Blue's approach was based on explicit programming and search, whereas later systems like Google DeepMind's AlphaZero used self-play and Neural network training to achieve superhuman chess performance without human-derived heuristics.
Legacy
The 1997 match is frequently referenced in discussions of human-computer interaction and the trajectory of AI development. It demonstrated that specialized hardware and algorithmic search could outperform a human world champion in a complex domain, but it also highlighted the limitations of narrow AI. The match remains a popular case study in computer science education and is often compared to later AI achievements such as Waymo's self-driving cars and OpenAI's language models.
Kasparov later wrote a book about the experience, "Deep Thinking," and became an advocate for human-machine collaboration, coining the term "advanced chess" where humans and computers play together. IBM's team received the Fredkin Prize for the achievement, and the match is commemorated in various technology museums.
Controversies and Criticisms
Critics noted that Deep Blue's victory was partly due to its enormous computational speed rather than general intelligence. The system could not learn from its games and had no understanding of chess beyond its programmed evaluation. Some also questioned the fairness of the match conditions, as Kasparov had not had access to Deep Blue's game logs from the 1996 match, while IBM had analyzed all of Kasparov's published games.
Despite these criticisms, the match is universally recognized as a turning point. It showed that with sufficient engineering, computers could match or exceed human performance in specific intellectual tasks, paving the way for later developments in Artificial intelligence research and applications.