Hacker News new | ask | show | jobs
Mapping GPUs to LLMs (and back): A bandwidth-based estimator for local inference (localllm-advisor.com)
2 points by apignotti 64 days ago