[morkfromork]

Imagine the type of machine learning we could do on low power 384GHz multi-cores

[aheilbut]

Imagine the type of machine learning we could do on 384 low power 1GHz cores.

[pjscott]

Not everything parallelizes nicely. Given enough processors, serial computation eventually becomes the bottleneck:

http://en.wikipedia.org/wiki/Amdahl's_law

[aheilbut]

Many machine learning methods parallelize extremely nicely.

[ioquatix]

Imagine the type of machine learning we could do on low power 384GHz quantum multi-cores with plasma exhaust relays and phase lock inverters.

[djillionsmix]

Invert the polarity

[smsm42]

I'm afraid the machine you'd get would be a furnace or a torch... AFAIK power that you'd need to dissipate is linear on frequency, so unless we will discover technology to radically lower the voltage used in CPUs or some other parameter, dissipating 100x heat would be kind of a big problem for such CPU.