AMD Offers Day 0 Support for Alibaba Qwen3.8 27B Model

AMD enables immediate local execution of Alibaba's Qwen3.8 27B on its hardware, achieving up to 51.8 tokens per second on Radeon GPUs.

Woofun AI reports that AMD has confirmed "Day 0" support for Qwen3.8 27B, the latest open-source model from Alibaba. This integration allows developers to run the 27-billion-parameter model locally on AMD processors and GPUs immediately upon release. The model is optimized for code generation, research, and long-context applications, running on AI PCs or single GPUs with 32GB memory. Preliminary tests indicate inference speeds of up to 24.5 tokens per second on Ryzen AI Max+ 395 processors and 51.8 tokens per second on Radeon AI PRO R9700 GPUs using llama.cpp with MTP optimization.