This is probably referencing the same idea that seeded https://cstheory.stackexchange.com/questions/12905/speedup-from-algorithmic-advances-vs-hardware https://

This is probably referencing the same idea that seeded https://cstheory.stackexchange.com/questions/12905/speedup-from-algorithmic-advances-vs-hardware https://cstheory.stackexchange.com/questions/12905/speedup-f.... I remember the claim, as well. That stack exchange indicates that it was never actually quantified. My gut would be that some of the algorithm advances require far more space that was available on early machines. Still a neat idea to look into.