3 ms·
Entirely due to the unified RAM between CPU and GPU in Apple Silicon. Laptops otherwise almost never have a GPU with sufficient RAM for LLMs.
by telotortium 1y ago
Entirely due to the unified RAM between CPU and GPU in Apple Silicon. Laptops otherwise almost never have a GPU with sufficient RAM for LLMs.
- rickydroll 1y agoShould have been clearer. I was thinking of a dedicated in-house LLM server I could use from different laptops.