AMD Announces Day 0 Local Support for Qwen3.8 27B Model

According to Deep Tide TechFlow on August 14, AMD announced via its official blog that starting today, it will provide Day 0 support for the Qwen3.8 27B model. Users can locally run the model using AMD Ryzen™ AI Max+ processors or AMD Radeon™ AI PRO R9700 32GB graphics cards, leveraging open-source frameworks like llama.cpp. Initial tests show speeds of up to 24.5 tokens/s on the Ryzen AI Max+ 395 and up to 51.8 tokens/s on the Radeon AI PRO R9700. Additionally, AMD recommends users quickly experience the model through LM Studio and integrate it into applications for local AI inference deployment using AMD's local development platform, Lemonade.