This isn't really feasible due to the limited propagation speed of em-waves (ie. electronic signals). I mean, sure, it's probably possible to get to, say, 50 or

This isn't really feasible due to the limited propagation speed of em-waves (ie. electronic signals). I mean, sure, it's probably possible to get to, say, 50 or 100 GHz with very deep pipelines (ie. only limited propagation of signals in each cycle), but we've already seen 15 years ago that deep pipelines basically don't work for general purpose computing. Additionally the frequency-dependent losses are unlikely to fundamentally change.