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

Nice experiment. Also a reminder that while you can buy a speed up with better hardware looking for a better algorithm usually gives higher rewards.