Windows Copilot will receive RTX GPU support – Nvidia Confirms

GPU support is coming to Copilot, allowing it to utilise the power of GeForce RTX

When Microsoft revealed their first Windows Copilot+ laptops, all of them used Qualcomm’s new Snapdragon X Elite SOC. None of them were powered using the AI performance of AMD or Intel NPUs, or by the AI power of GeForce RTX GPUs. At Computex, Nvidia has confirmed that they are working with Microsoft to bring GPU support to the Copilot Runtime later this year.

With Nvidia being the largest provider of AI-enhanced PC hardware, it makes sense for Microsoft to support their graphics cards. Nvidia has shipped GPUs with AI accelerators since the launch of their RTX 20 series in 2018. Beyond that, their latest GPUs deliver AI performance levels that are much higher than today’s NPUs. There’s a reason why Nvidia says that NPUs are for “light AI” workloads.

Windows Copilot Runtime to Add GPU Acceleration for Local PC SLMs
Microsoft and NVIDIA are collaborating to help developers bring new generative AI capabilities to their Windows native and web apps. The collaboration will provide
application developers with easy application programming interface (API) access to GPU accelerated small language models (SLMs) that enable retrieval-augmented generation (RAG) capabilities that run on-device powered by Windows Copilot Runtime.

While NPUs will be bringing AI performance to the latest processors from AMD, Intel, and Qualcomm, Nvidia sees themselves as the place where “heavy AI” happens. Nvidia has offered more AI TOPS than is required for Copilot for years. That said, Microsoft has not allowed their Copilot runtime to access this performance. Well, at least not yet.

At least initially, Microsoft appears to be focused on getting Copilot to run on the NPUs within modern processors. After that, GPU support will arrive to give users access to the boosted AI performance of discrete GPUs.

You can join the discussion on Copilot getting support for RTX on the OC3D Forums.