§ feed · storyline
AMD's vLLM-ATOM plugin boosts LLM inference on MI350 and MI400
AMD releases the vLLM-ATOM plugin to accelerate LLM inference for models including DeepSeek-R1 and Kimi-K2 on Instinct MI350 and MI400 accelerators.
AMD’s vLLM-ATOM Plugin Supercharges DeepSeek-R1, Kimi-K2, and gpt-oss-120B AI LLM Inference on Instinct MI350 and MI400 Accelerators Wccftech
§ sources1 publication · timeline below