[bjackman]

It doesn't matter if you have the RAM, running a 1.5TB model for a single context stream is fundamentally inefficient.

[Tepix]

Lots of things are inefficient in IT, yet we do them anyway. Think about the amount of time your laptop is idle.