Though untimely optimization generally is a actually dangerous factor, late optimization is typically unavoidable as a result of we merely didn’t have find out about sure algorithms but to massively enhance the fixing of an issue. When the Nineteen Nineties was all about beating chess grandmasters at their recreation, IBM went as far as to construct a the Deep Blue customized tremendous laptop for it. That is what [Nicholas Carlini] managed to successfully beat utilizing a model of the Stockfish chess engine operating on an era-appropriate Pentium PC.
A standard approach to measure one’s talent in chess includes the Elo score system, named after chess grasp and physics professor [Arpad Elo]. As a relative system, it may be used to match completely different gamers together with chess engines towards one another, with IBM’s Deep Blue scoring 2700, with the modern-day Stockfish engine hitting round 3700. This makes Stockfish statistically more likely to win towards a human chess grandmaster than Deep Blue, but in addition raises the query of how a lot of this benefit is because of Stockfish operating on a a lot quicker system.
To analyze the influence right here, [Nicholas] used a Pentium Professional 200 MHz system with 64 MB of RAM (Quick Web page or EDO SIMMs), which might have been a quite good system for about 1995. Two chess engines have been ported to this platform, being Artful and Stockfish. Subsequent each engines have been tweaked with the assistance of an LLM coding bot to see what distinction varied compilation flags and algorithms would make.
Based mostly on the outcomes it might appear that the reply is ‘sure’: this 200 MHz Pentium Professional might have probably crushed even a chess grandmaster like [Kasparov] and Deep Blue in at the very least just a few video games, although with out going again in time by about thirty years we’ll by no means know for sure.
Header: Wilfredor, CC0
