[zhye]

Serving LLM with AMD GPUs to serve LLM looks impressive, MLC is evolving fast! Any results on NVLink/xGMI instead of PCIe?