On November 12th， AMD unveiled its latest Zen architecture roadmap at the Financial Analyst Day 2025 (FAD 2025)， officially confirming development cycles for two core architectures， Zen 6 and Zen 7， and shedding light on their key technological directions. This announcement signifies AMD’s strategic push into advanced manufacturing processes and a significant emphasis on artificial intelligence capabilities.
Zen 6 is slated for release in 2026， and will be the first to leverage TSMC’s 2nm process technology. The primary focus for Zen 6 will be on enhancing both performance and efficiency， building upon the established strengths of the Zen architecture.
A more significant development is expected with Zen 7. According to the roadmap， Zen 7 will utilize a “Future Node” process technology. While not yet explicitly named， industry speculation suggests this will fall between TSMC’s N2P and N1.8 manufacturing processes， indicating a continued drive towards miniaturization and improved transistor capabilities.
The most groundbreaking aspect of Zen 7 lies in its introduction of a new Matrix Engine (for matrix operations) and enhanced AI Data Format support， alongside a reinforced AI Pipeline structure. These upgrades are poised to make Zen 7 the first x86 architecture processor to feature dedicated AI acceleration engines， marking a pivotal entry for the Zen platform into the era of AI-native computing.
This strategic shift clearly indicates that AMD is no longer solely focused on performance and power efficiency as its primary competitive differentiators. Given Intel’s current competitive landscape， where AMD has established a strong lead， the company is now prioritizing AI acceleration as a core element of its next-generation CPUs. This foresight positions AMD to capitalize on the rapidly growing AI market.
This strategy is not limited to the PC and server markets alone; it also signals AMD’s intention to play a crucial role in the evolving AIPC (AI Personal Computer) ecosystem. The integration of dedicated AI hardware within mainstream processors will enable more efficient and powerful on-device AI processing， a key driver for next-generation computing experiences.
Based on current market projections， Zen 7 is anticipated to launch between 2027 and 2028. It is expected that the Verano series of EPYC data center CPUs will be among the first products to adopt the Zen 7 architecture， underscoring the significance of AI acceleration for enterprise-level computing and data processing.
