public void Dispatch (int kernelIndex, int threadGroupsX, int threadGroupsY, int threadGroupsZ);


kernelIndex Which kernel to execute. A single compute shader asset can have multiple kernel entry points.
threadGroupsX Number of work groups in the X dimension.
threadGroupsY Number of work groups in the Y dimension.
threadGroupsZ Number of work groups in the Z dimension.


Execute a compute shader.

This functions "runs" the compute shader, launching the indicated number of compute shader thread groups in the X, Y and Z dimensions. Within each work group, a number of shader invocations ("threads") are made. The work group size is specified in the compute shader itself (using "numthreads" HLSL attribute), and the total amount of compute shader invocations is thus group count multiplied by the thread group size. Work group size can be queried using GetKernelThreadGroupSizes function.

See Also: FindKernel, GetKernelThreadGroupSizes, DispatchIndirect, Compute Shaders.