GPU

Experimental: This is an experimental technology
Check the Browser compatibility table carefully before using this in production.

The GPU interface of the WebGPU API is the starting point for using WebGPU. It can be used to return a GPUAdapter from which you can request devices, configure features and limits, and more.

The GPU object for the current context is accessed via the Navigator.gpu or WorkerNavigator.gpu properties.

Instance properties

wgslLanguageFeatures Experimental Read only

A WGSLLanguageFeatures object that reports the WGSL language extensions supported by the WebGPU implementation.

Instance methods

requestAdapter() Experimental

Returns a Promise that fulfills with a GPUAdapter object instance. From this you can request a GPUDevice, which is the primary interface for using WebGPU functionality.

getPreferredCanvasFormat() Experimental

Returns the optimal canvas texture format for displaying 8-bit depth, standard dynamic range content on the current system.

Examples

Requesting an adapter and a device

js

async function init() {
  if (!navigator.gpu) {
    throw Error("WebGPU not supported.");
  }

  const adapter = await navigator.gpu.requestAdapter();
  if (!adapter) {
    throw Error("Couldn't request WebGPU adapter.");
  }

  const device = await adapter.requestDevice();

  //...
}

Configuring a GPUCanvasContext with the optimal texture format

js

const canvas = document.querySelector("#gpuCanvas");
const context = canvas.getContext("webgpu");

context.configure({
  device: device,
  format: navigator.gpu.getPreferredCanvasFormat(),
  alphaMode: "premultiplied",
});

Specifications

Specification
WebGPU
# gpu-interface

Browser compatibility

BCD tables only load in the browser

See also