Tag
GLM 5.2, an open frontier-scale AI model, is now available on Microsoft Foundry with AMD MI300X hardware, enabling efficient Codex goal execution.
This blog post details the author's efforts to get DeepSeek-V4-Flash running on AMD MI300X GPUs, highlighting software compatibility issues with the FP8 dialect and providing a worklog of the process.
Kog AI achieves 3,000 tokens/s inference speed on 8× AMD MI300X GPUs and 2,100 on 8× NVIDIA H200, leveraging a hidden efficiency gap in GPU token generation.
A monokernel approach for LLM decoding on AMD MI300X GPUs achieves up to 3,300 output tokens/s per request without speculative decoding or quantization, using memory access patterns mapped to the die topology.
Shares early benchmark scores and evaluation metrics for an open-weight model stack run on a single AMD MI300X, noting competitive performance against closed-source alternatives.