Could a modern chess engine on a Pentium have defeated Kasparov in 1996?

The central claim is that Deep Blue’s specialized hardware was not necessary to beat Garry Kasparov: modern chess-engine methods can make a thirty-year-old desktop extraordinarily strong. Deep Blue combined 30 processors with 480 chess-specific chips and searched more than 200 million positions per second. Yet most of the roughly 1,335-Elo rise from 1996-era engines to today’s strongest engines came from better programming, not faster computers. The experiments suggest that improved search, move selection, evaluation, and tuning account for about 720 Elo of that gain.

To test the idea, the author adapted two engines to a 200 MHz Pentium Pro with 64 MB of RAM. Updating the 1996 engine Crafty with modern techniques raised its estimated strength to 2,718 Elo at one second per move and 2,974 at ten seconds. A Stockfish port, further adjusted for the old processor, reached 2,846 and 3,109 Elo at those time controls, and 3,180 at 30 seconds. Its estimated classical strength of about 3,300 is extrapolated from shorter games, not directly measured; the available match data are also limited. Still, in 100 games, the Pentium-based modern engine beat Crafty running on modern hardware 78 times, drew 17, and lost five. The estimates imply it would have been a heavy favorite against Kasparov, though the classical rating is uncertain.