Interface GPUComputePassEncoder

The GPUComputePassEncoder interface of the WebGPU API encodes commands related to controlling the compute shader stage, as issued by a GPUComputePipeline. It forms part of the overall encoding activity of a GPUCommandEncoder. Available only in secure contexts.

MDN Reference

interface GPUComputePassEncoder {
    label: string;
    dispatchWorkgroups(
        workgroupCountX: number,
        workgroupCountY?: number,
        workgroupCountZ?: number,
    ): void;
    dispatchWorkgroupsIndirect(
        indirectBuffer: GPUBuffer,
        indirectOffset: number,
    ): void;
    end(): void;
    insertDebugMarker(markerLabel: string): void;
    popDebugGroup(): void;
    pushDebugGroup(groupLabel: string): void;
    setBindGroup(
        index: number,
        bindGroup: GPUBindGroup | null,
        dynamicOffsets?: number[],
    ): void;
    setBindGroup(
        index: number,
        bindGroup: GPUBindGroup | null,
        dynamicOffsetsData: Uint32Array<ArrayBufferLike>,
        dynamicOffsetsDataStart: number,
        dynamicOffsetsDataLength: number,
    ): void;
    setBindGroup(
        index: number,
        bindGroup: GPUBindGroup | null,
        dynamicOffsets?: Iterable<number, any, any>,
    ): void;
    setPipeline(pipeline: GPUComputePipeline): void;
}

Hierarchy (View Summary)

Index
label
dispatchWorkgroups dispatchWorkgroupsIndirect end insertDebugMarker popDebugGroup pushDebugGroup setBindGroup setPipeline
label: string

MDN Reference

  • The dispatchWorkgroups() method of the GPUComputePassEncoder interface dispatches a specific grid of workgroups to perform the work being done by the current GPUComputePipeline (i.e., set via GPUComputePassEncoder.setPipeline()).

    MDN Reference

    Parameters

    • workgroupCountX: number
    • OptionalworkgroupCountY: number
    • OptionalworkgroupCountZ: number

    Returns void

  • The dispatchWorkgroupsIndirect() method of the GPUComputePassEncoder interface dispatches a grid of workgroups, defined by the parameters of a GPUBuffer, to perform the work being done by the current GPUComputePipeline (i.e., set via GPUComputePassEncoder.setPipeline()).

    MDN Reference

    Parameters

    • indirectBuffer: GPUBuffer
    • indirectOffset: number

    Returns void

  • The end() method of the GPUComputePassEncoder interface completes recording of the current compute pass command sequence.

    MDN Reference

    Returns void

Was this helpful?