The promise of the “AI PC” felt revolutionary just months ago. Microsoft declared every Windows 11 machine would possess AI capabilities, yet the fervor surrounding Copilot+ PCs at CES 2026 felt…off. It was as if the industry was scrambling to meet a goal Microsoft had quietly begun to move away from.
A wave of Copilot+ laptops dominated the show floor, but a nagging question lingered: were these machines chasing a future Microsoft had already redefined? The initial emphasis on dedicated Neural Processing Units (NPUs) now feels muted, especially with Microsoft’s new Windows AI Foundry prioritizing GPUs and CPUs for AI tasks – a shift from the original Copilot Runtime’s NPU reliance.
NPUs are appearing in more and more devices, and finally achieving the processing power initially demanded. When Microsoft first introduced Copilot+ PCs, a minimum of 40 trillion operations per second (TOPS) was the benchmark. This presented a significant hurdle for Intel, whose processors typically offered only 13 TOPS, with Lunar Lake reaching 48 TOPS. For much of 2025, many reviewed laptops simply couldn’t meet the standard.
At CES 2026, Intel showcased its Core Ultra Series 3 (Panther Lake) processors, boasting a 50 TOPS NPU alongside improved multithreaded performance. This represents a major leap forward, addressing the previous limitations of Lunar Lake, which prioritized battery life over sustained performance. The race to deliver capable NPUs is officially on.
AMD isn’t far behind, with its Ryzen AI 400 series featuring a 60 TOPS NPU for both laptops and desktops. While an improvement over the previous generation, the practical impact of this 10 TOPS increase remains to be seen, as few applications currently leverage NPU power. Qualcomm continues to lead in TOPS performance with its Snapdragon X platforms, but the core issue persists: a compelling reason to *need* an NPU is still elusive.
Microsoft initially positioned Copilot+ PCs as the pinnacle of AI integration, but then announced a broader strategy: bringing AI capabilities to *all* Windows 11 PCs. Yusuf Mehdi, a Microsoft executive, explained the shift, emphasizing the desire to make AI accessible to everyone, not just those with specialized hardware. This change in direction casts a shadow over the recent NPU push.
The reality is, the exclusive features offered by Copilot+ PCs – enhanced webcam effects, image generation, PC usage search, and basic text actions – haven’t sparked a revolution. A recent conversation with a rideshare driver highlighted this disconnect. He was unaware of the Copilot+ distinction, and the benefits simply didn’t resonate with his everyday Windows 11 experience.
Looking ahead, it seems increasingly likely that Microsoft will prioritize GPU and CPU-powered AI features for all Windows 11 users. A future update could unlock Copilot+ capabilities on a wider range of hardware, effectively bypassing the NPU requirement altogether. This would be a more inclusive and impactful approach.
Perhaps Microsoft prematurely focused on NPUs, creating a hardware race before the software ecosystem was ready. Even a high-end gaming PC, equipped with a powerful discrete GPU, remains ineligible for Copilot+ features – a baffling situation that underscores the current imbalance between hardware and software.
