AMD is expanding its vision for personal AI with client platforms designed to bring more intelligence closer to where work happens.
As the rapid adoption of agentic AI increases demand for local compute, AMD Ryzen AI systems are helping developers build, test and run capable models on personal devices.
At Advancing AI 2026 this week, AMD demonstrated how with Ryzen AI Halo, an expanded Hugging Face partnership, and a new collaboration with Cisco, AMD is advancing an open client AI ecosystem built for performance, flexibility and scale.
Agentic AI is moving beyond single prompts to systems that reason, use tools, coordinate with other agents, and complete work continuously. That shift increases demand across the computing ecosystem, making client systems an important complement to the data center. Running more AI locally can help improve privacy, responsiveness, reliability and efficiency while allowing the cloud to focus on the largest frontier workloads.
“AI is moving from a cloud-only experience to something deeply personal, local and always available,” says Jack Huynh, senior vice-president and GM: Computing and Graphics Group at AMD. “With Ryzen AI platforms, AMD is helping developers, enterprises and creators bring powerful AI closer to where work happens — unlocking more responsive, private and efficient experiences on the client.”
The PC is already one of the most powerful compute engines in the hands of an individual. Across billions of devices, enormous compute capacity is already deployed, but much of its AI potential remains untapped. As more workloads run on systems already in place, users can benefit from lower latency, greater control over data and more resilient experiences when connectivity is limited.
Smaller, local AI models are now similar in quality to frontier models, changing the compute required to deliver advanced intelligence. Workloads once reserved for large-scale infrastructure can increasingly run on powerful personal systems. AMD Ryzen AI platforms are designed for this shift, enabling users to run capable models locally while preserving headroom for more sophisticated agentic workflows.
Ryzen AI Halo gives developers a local platform for building, testing and iterating on AI models, with 128Gb of unified memory and support for models up to 200-billion parameters.
AMD is also expanding its partnership with Hugging Face to bring optimized open models, libraries and toolkits to Ryzen AI Halo, helping developers spend less time configuring infrastructure and more time building applications.
And, later this year, every AMD Ryzen AI Halo developer platform will ship with a free year of Hugging Face PRO.