Synopsys frames physical AI as a systems design challenge

AI is moving beyond the data center into cars, factories, wearables, humanoid robots, and service robots, creating what Synopsys describes as the era of physical AI. Cloud AI continues to train models and run large-scale inference, while edge AI puts models on devices for low latency, real-time response, privacy, and resilience when disconnected.

Edge and physical AI products are projected to grow at roughly 35% CAGR through 2035, driven by sensing advances, humanoid and collaborative robots, automation, and on-device AI. The shift turns chip development into a broader systems problem, with teams balancing deterministic real-time behavior, heterogeneous compute efficiency, functional safety, power, cost, and time-to-market pressure.

Synopsys is positioning a silicon-to-systems approach that combines process technology, silicon IP, and design methodology earlier in development. Its portfolio spans high-speed interfaces, foundation IP, and silicon lifecycle management components designed for deterministic performance, low power, cost discipline, and functional safety.

The company also highlights work with TSMC’s compact node platforms N6C (also known as N6 V1.1) and N4C, described as cost-optimized variants of N6 and N4P for high-volume AI designs. The goal is to give engineering teams a predictable foundation so they can focus on product-level AI capability, applications, and user experience.