GPUBuffer
WebGPU API
WebGPU API
The GPUBuffer interface of the WebGPU API represents a block of memory that can be used to store raw data to use in GPU operations.
A GPUBuffer object instance is created using the GPUDevice.createBuffer() method.
Instance properties
- label
- : A string providing a label that can be used to identify the object, for example in GPUError messages or console warnings.
- mapState
- : An enumerated value representing the mapped state of the
GPUBuffer. - size
- : A number representing the length of the
GPUBuffer's memory allocation, in bytes. - usage
- : The bitwise flags representing the allowed usages of the
GPUBuffer.
Instance methods
- destroy()
- : Destroys the
GPUBuffer. - getMappedRange()
- : Returns an ArrayBuffer containing the mapped contents of the
GPUBufferin the specified range. - mapAsync()
- : Maps the specified range of the
GPUBuffer. Returns a Promise that resolves when theGPUBuffer's content is ready to be accessed with GPUBuffer.getMappedRange(). - unmap()
- : Unmaps the mapped range of the
GPUBuffer, making its contents available for use by the GPU again.
Examples
In our basic compute demo, we create an output buffer to read GPU calculations to, and a staging buffer to be mapped for JavaScript access.
const output = device.createBuffer({
size: BUFFER_SIZE,
usage: GPUBufferUsage.STORAGE | GPUBufferUsage.COPY_SRC,
});
const stagingBuffer = device.createBuffer({
size: BUFFER_SIZE,
usage: GPUBufferUsage.MAP_READ | GPUBufferUsage.COPY_DST,
});
Later on, once the stagingBuffer contains the results of the GPU computation, a combination of GPUBuffer methods are used to read the data back to JavaScript so that it can then be logged to the console:
- GPUBuffer.mapAsync() is used to map the
GPUBufferfor reading. - GPUBuffer.getMappedRange() is used to return an ArrayBuffer containing the
GPUBuffer's contents. - GPUBuffer.unmap() is used to unmap the
GPUBufferagain, once we have read the content into JavaScript as needed.
// map staging buffer to read results back to JS
await stagingBuffer.mapAsync(
GPUMapMode.READ,
0, // Offset
BUFFER_SIZE, // Length
);
const copyArrayBuffer = stagingBuffer.getMappedRange(0, BUFFER_SIZE);
const data = copyArrayBuffer.slice(0);
stagingBuffer.unmap();
console.log(new Float32Array(data));
Specifications
Browser compatibility
See also
- The WebGPU API