Hacker News new | past | comments | ask | show | jobs | submit
zzzoom | on: Making Deep Learning Go Brrrr from First Principles (2022)
EPYC 9965: 614GBps of 12-channel DDR5-6400

A100: 1935GBps of HBM2e

Most of those FLOPS are constrained by memory bandwidth.

loading story #48251822