NVDA N1X CPU Targets 150 Million Annual PC Units at Computex June 1

NVIDIA's N1X CPU reveal at Computex on June 1 is shaping up as a major inflection point for the PC market. The chip integrates both CUDA cores and RTX 5070 GPU on a single die, combining what have historically been separate processor and graphics functions. This is not merely an incremental refresh but a fundamental rearchitecting of client computing around AI-native workloads at the edge.

The addressable market is substantial: Windows PCs span roughly 150-270 million units annually, valued at approximately $50-65 billion. Intel and AMD have long dominated this space through x86 licensing, but NVIDIA's integrated approach sidesteps many of their cost and power advantages. Early commentary from traders suggests the N1X is a 'legacy chip killer,' potentially displacing both AMD's Ryzen and Intel's Core architectures in performance-per-watt for agentic AI and local inference workloads.

Jensen Huang's keynote at GTC Taipei on Sunday night (June 1) is expected to position this as a 'new era of PC computing,' framing the transition from cloud-dependent AI to on-device agentic systems. The timing aligns with broader softening in semiconductor valuations, meaning markets are repricing expectations for where the next layer of AI capex lands.

Skeptics note that ARM-based Windows adoption has failed repeatedly since 2016-2020, and echoes of those failed ARM Windows initiatives linger. Yet the presence of proprietary CUDA on the N1X, plus NVIDIA's installed base of developers, creates a lock-in dynamic that prior ARM attempts lacked. Market reaction will hinge on real-world performance benchmarks and OEM adoption announcements at or shortly after Computex.