Comment

Comment by charlescurt123

comparing a laptop to a A100 (312 teraFLOPS) or H100 (~1P FLOPS) server is a stretch to say the least.

An M2 is according to a reddit post around 27 tflops

So < 1/10 the performance of just computation. let alone the memory.

What workflow would use something like this?

Discussion

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton