Microsoft DirectML now supports Copilot+ PCs and WebNN

DirectML is a low-level API that enables developers to run machine learning workloads on any DirectX 12 compatible GPUs from AMD, Intel, and NVIDIA. It was first introduced in Windows 10, version 1903. Recently, Microsoft started supporting NPUs found in modern SoCs in DirectML. Today, Microsoft announced that DirectML now supports Copilot+ PCs, powered by Qualcomm Hexagon NPU in the Snapdragon X Elite Compute Platform.

DirectML on Copilot+ PCs will be supported with the following system requirements:

DirectML minimum version of 1.15.2 (ARM64)
ONNX Runtime minimum version of 1.18 (ARM64)
Windows 11, version 24H2 or newer
Qualcomm Hexagon NPU Driver minimum version 30.0.31.250 or newer

Upendra Kulkarni, Vice President, Compute Software Product Management at Qualcomm, said:

“With Snapdragon X Elite, we introduced an industry-leading NPU with 45 TOPS of AI performance at incredible power efficiency. DirectML is a developer-friendly ML programming interface that uses a familiar DirectX API structure. By supporting DirectML on our NPU, developers are now able to easily access its phenomenal capability and can port their models from GPU to NPU with minimal effort. We collaborated extensively with Microsoft to optimize DirectML for NPU to maximize hardware performance. We are excited to be co-announcing this developer preview program.”

Along with Copilot+ PCs support, Microsoft today announced that web developers can now take advantage of NPUs with the latest release of DirectML and WebNN Developer Preview.

WebNN is a new web standard that allows developers to run machine learning models in the browser for AI-based web applications. This new DirectML update adds support for an initial set of models on Intel Core Ultra processors with Intel AI Boost and Copilot+ PCs powered by Qualcomm Hexagon NPUs. Developers can now try it out today on the WebNN Developer Preview website.

WebNN on DirectML will be supported with the following system requirements: