The OpenXLA compiler stack and JAX now run upstream on AMD ROCm, with XLA gating every pull request on real AMD Instinct silicon through GitHub Actions and JAX running hardware tests on every ROCm PR.…
AMD introduced workload pre-emption in its Resource Manager to reclaim idle GPUs from underutilized workloads, improving cluster efficiency. The feature monitors GPU utilization per workload and termi…
AMD introduces DP Attention and Two-Batch Overlap (TBO) optimizations for DeepSeek-V4 inference on MI355X GPUs, using a coordinated prefill scheduler called PrefillDelayer to reduce padding waste and …
AMD Radeon GPUs, both integrated and discrete, now support running large language models locally through open-source tools like Lemonade, LM Studio, Ollama, and llama.cpp. A new guide provides step-by…
A comparative analysis of scale-out RoCE network traffic patterns in training large language models reveals that hyperscale deployments like xAI's Grok 4.0 generate network loads orders of magnitude h…