AMD Advancing AI 2026 — Building What's Next in AI

Advancing AI 2026 is AMD's annual developer, customer and partner conference, returning to San Francisco's Moscone West on July 22-23 and livestreamed free on AMD's channel. It is the company's headline moment for laying out its end-to-end AI strategy, from silicon to software, and a barometer for the wider AI-hardware race. The agenda is expected to spotlight next-generation Instinct accelerators, rack-scale "AI factory" systems, and the open ROCm software stack — AMD's pitch for an open alternative to proprietary GPU ecosystems as enterprises move from AI experimentation to production deployment.

For attendees, the draw is two days of technical depth rather than keynote theater: hands-on developer workshops, deep dives on inference and agentic workloads, and "meet the experts" sessions on the unglamorous-but-decisive problems of scaling AI — network transport between GPU clusters, congestion control, and matching workloads to the right compute. Founders building on AMD infrastructure are featured directly, reflecting how fast independent labs and digital natives are shipping production AI on non-Nvidia stacks. With AI compute supply and cost now the gating factor for most frontier and enterprise deployments, Advancing AI is one of the most consequential vendor conferences on the 2026 calendar for anyone building, buying or deploying AI infrastructure at scale.