Microsoft's Windows ML platform is now generally available for production purposes, making it easier for developers to add AI to Windows apps. Microsoft says that Microsoft is supported on all devices running Windows ML on Windows 11 24H2 and later, and provides a local AI inference framework that helps you build a more “responsive, private, and cost-effective” AI experience.
“The way this works is for silicon partners to build and maintain a running provider that serves as a way for Windows ML to distribute, manage and register to run AI workloads, run devices on performance, act as a layer of hardware abstraction for developers, and get optimal performance for each particular silicon,” Microsoft said in the announcement. Simply put, the platform allows AI-injected apps to take advantage of the best PC hardware for a particular workload, such as GPUs for power-intensive tasks, NPUs for power efficiency, CPUs for flexibility, and more.
Windows ML has been in development for several years and was first announced in 2018 on Microsoft's Windows 10 operating system. Microsoft says software developers, including Adobe, McAfee and Topaz Labs, are working on adopting Windows ML in future releases after helping them test the platform during development.
Future integrations of Adobe's Premiere Pro and After Effects Video Editing Software will leverage local NPU for semantic search, audio tagging and scene editing detection capabilities. McAfee also uses Windows ML to automatically detect Deepfake videos and fraud on social networks, and Topaz Labs has leveraged the platform to develop AI-powered image editing capabilities with Topaz photos. With Windows ML streamlining AI development with Windows-based apps, Microsoft wants an influx of similar AI experiences that can attract more consumers to the operating system.
