Skip to main content

Overview

The @luma.gl/core module provides an abstract API that enables application code to portably work with both WebGPU and WebGL. The main export is the Device class which provides methods for creating GPU resources such as Buffer, Texture, Shader etc.

Installing adapters​

The @luma.gl/core module is not usable on its own. A device adapter module must be imported and registered.

import {luma} from '@luma.gl/core';
import {WebGPUAdapter} from '@luma.gl/webgpu';

luma.registerDevice([WebGPUAdapter])
const device = await luma.createDevice({type: 'webgpu', canvas: ...});

It is possible to register more than one device adapter to create an application that can work in both WebGL and WebGPU environments.

luma.registerDevice([WebGPUAdapter])
import {luma} from '@luma.gl/core';
import {WebGPUAdapter} from '@luma.gl/webgpu';
import {WebGLAdapter} '@luma.gl/webgl';

const webgpuDevice = luma.createDevice({type: 'best-available', canvas: ...});

Creating GPU Resources​

Once the application has created a Device, GPU resources can be created:

const buffer = device.createBuffer(...)

Accessing the CanvasContext​

A Device may (optinally) be used to render in one or more canvases (HTML canvas elements). The connection between a Device and a canvas is managed by the CanvasContext class.

info

In WebGL there is always exactly one canvas associated with the device and it is not possible to create a canvas-less context or render into multiple contexts.