Comment by nox101

with all of the local ML being introduced by Apple and Google and Microsoft this thinking seems close to "640k is all you need"

I suspect consumer workloads to rise