Exciting news for Web AI developers! We are thrilled to announce the result of a partnership between Intel and Microsoft to bring the WebNN API to the Microsoft Edge browser as a developer preview feature on Windows. Meanwhile, Intel has been focusing on contributing to the WebNN implementation in Google hosted Chromium.org
What is WebNN?
WebNN stands for web neural network. This innovative API empowers web developers to harness the power of their users' devices for machine learning tasks. Through WebNN, developers can leverage the AI PC Artificial Intelligence processing cores and accelerators, including CPUs, GPUs, and NPUs, in a unified manner. This translates to power-efficient, faster processing and smoother performance for web applications that utilize machine learning models.
Benefits for Developers
The WebNN API in the browser offers several advantages for developers:
- Unified Access to the AI PC Accelerators: No more wrestling with device-specific code! WebNN provides a consistent interface for utilizing various AI PC accelerators, simplifying development, and ensuring broader compatibility.
- Enhanced Performance: By tapping into the power of AI PC accelerators, web apps can achieve significant performance gains and power efficiency, especially for machine learning workloads.
- Unlocking New Possibilities: WebNN opens doors for creating a new generation of intelligent web applications that leverage on-device machine learning for tasks like image recognition, natural language processing, and more.
Get Started Today
Ready to dive in and explore the potential of WebNN? The preview feature is now available in the latest Canary version of the Microsoft Edge browser. Download the update and start experimenting with WebNN to deliver exceptional AI user experiences.
You can get the instructions to enable the WebNN capability on this page; visit the WebNN Developer Preview site to check out the latest WebNN demo cases. We encourage developers to explore the WebNN API documentation and join the conversation online to share your feedback and experiences. Together, let's push the boundaries of what is possible on the open web platform with machine learning!
About the Author
Qi Zhang, Director of Web Platform Engineering, Intel China
Qi Zhang is the director of the web platform engineering team, which is part of Intel China’s System Software Engineering Group. Qi’s team works to enable and optimize web technology for Intel platforms, focusing on key runtime components of Chromium including V8, WebAssembly, Blink, and key web APIs such as WebGPU, WebRTC, and WebNN. Connect with him on LinkedIn.