Windows Copilot AI will eventually run on GeForce RTX GPUs

Microsoft, Intel, AMD, and Qualcomm’s Plans for AI PCs

The big picture: Microsoft, Intel, AMD, and Qualcomm have unveiled their plans to introduce “AI PCs” that utilize neural processing units (NPUs) for onboard generative AI. Nvidia, however, believes that its RTX graphics chips are better suited for this task. Microsoft and Nvidia are collaborating to enable Copilot to utilize RTX hardware.

Developers creating Windows apps that leverage generative AI will soon have access to an API for running GPU-accelerated small language models through the Windows Copilot Runtime. This collaboration will allow Copilot and other generative AI applications to run on both NPUs and RTX GPUs, potentially expanding locally processed AI workloads on PCs.

At Computex this year, the focus has been on AI PCs, with Intel, AMD, and Qualcomm showcasing their upcoming processors. New chip lineups like Intel’s Lunar Lake, AMD’s Strix Point, and Qualcomm’s Snapdragon X emphasize how well their NPUs handle generative AI applications without relying solely on cloud servers. Apple is also expected to announce similar plans at WWDC.

Nvidia is pushing for its GPUs to be the preferred platform for client-side AI. The company claims that its entry-level and mid-range RTX 4000 graphics cards can process AI faster than the latest NPUs from CPU designers. RTX GPUs already have a wide user base, giving Nvidia an advantage in the AI PC market.

While current “RTX AI PC” applications are limited, enabling Copilot to utilize RTX hardware could expand the range of use cases and bring Microsoft’s AI assistant to more PCs. Nvidia has also introduced an AI-based gaming assistant similar to Copilot.

Nvidia showcased a series of “RTX AI” laptops from MSI and Asus, combining new AMD CPUs with the mobile RTX 4070. These devices are among the first to run Copilot on both NPUs and GPUs, with more Lunar Lake-based systems expected to follow suit.

Nvidia and Dell are hinting at a potential AI PC collaboration next year, possibly involving RTX GPUs and an Arm-based CPU or APU from Nvidia.