3 ms·VLLM X AMD: Efficient LLM Inference on AMD Instinct MI300X GPUs (Part 1)2 points by latchkey 2y ago