The Google Chrome team has announced the release of the WebGPU API for high-performance 3D graphics and data-parallel computation on the web. The new API is now available by default in Chrome 113, which is currently in the Beta Channel and will soon arrive to all users in the Stable Channel.
WebGPU is a new web standard and instruction set for hardware-accelerated graphics and computing developed by engineers from Google, Apple, Mozilla, Microsoft, and other companies. Like DirectX, Metal, and Vulkan, WebGPU allows rendering directly on a graphics processing unit (GPU), with access to more advanced features (unlike WebGL). WebGPU promises to reduce the JavaScript workload for the same graphics and more than 3x improvements in machine learning model interfaces.
Developers can draw images using WebGPU with highly-detailed scenes and many different objects (CAD models, for example), execute advanced algorithms for drawing realistic scenes, and run machine learning models more efficiently with better access to the GPU.
The new API is the result of more than six years of joint development (the initial design was developed in 2017). The first implementation is now available in Chrome, with Firefox and Safari to follow suit soon. WebGPU supports Windows PCs with DirectX 12, macOS, and ChromeOS with Vulkan support. Google says Android and Linux support will arrive shortly.
Developers can learn more about WebGPU support in Google Chrome in a blog post on the official Chrome Developers blog. It also includes links to additional resources, such as W3C specifications, best practices, MDN documentation, etc.