Dedicated neural net silicon is faster than GPU, at the cost of being less gener... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		dannyw on Dec 10, 2021 \| parent \| context \| favorite \| on: Apple Neural Engine Internal: From ML Algorithm to... Dedicated neural net silicon is faster than GPU, at the cost of being less general purpose. NVIDIA GPUs for example dedicate a modest fraction of their die to Tensor cores, at the cost of less general-purpose CUDA cores.

martinald on Dec 10, 2021 [–]

Thanks. I didn't realise how much faster. Probably you can do 10x the FLOPs in similar die space on dedicated silicion vs GPU. I thought it would be faster, but not an order of magnitude more.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact