> mostly driven my straightforward improvements in semiconductor technology
> (Moore's law, etc) and customer demand that wouldn't have helped bubble
> memory at all...


So, when will Moore's law break?
So far, it's held true, but we approach quantum level problems..
Any bets?