For anyone wondering about local LLM applications: - current price $200+ - limited VRAM (12-14 GB) - poor performamce in clusters [1] - only 2 lanes of PCI-e

For anyone wondering about local LLM applications: - current price $200+ - limited VRAM (12-14 GB) - poor performamce in clusters [1] - only 2 lanes of PCI-e 2.0 (still better than CMP XX HX mining cards) [1] https://ciroluciotecce.it/en/posts/bc250-cluster-inferenza-locale/ https://ciroluciotecce.it/en/posts/bc250-cluster-inferenza-l...