3D launch configuration for symmetry-reduced G-space loops (z dimension folded for real data sets) More...

#include <GpuKernelUtils.h>

Inheritance diagram for GpuLaunchConfigHalf3D:

Public Member Functions
template<typename GpuKernel >
	GpuLaunchConfigHalf3D (GpuKernel *gpuKernel, vector3< int > S)
	Just use the above after reducing the z-dimension to half.

Public Member Functions inherited from GpuLaunchConfig3D
template<typename GpuKernel >
	GpuLaunchConfig3D (GpuKernel *gpuKernel, vector3< int > S)
	Set up blocks and grid for a 1D operation over N data points.

Public Member Functions inherited from GpuLaunchConfig
template<typename GpuKernel >
	GpuLaunchConfig (GpuKernel *gpuKernel)
	Initialize the device and function properties.

Additional Inherited Members
Public Attributes inherited from GpuLaunchConfig3D
dim3	nPerBlock
	dimension of block

dim3	nBlocks
	dimension of grid (note nBlocks could be 3D for really large kernels)

int	zBlockMax
	Grids are 2D, so need to loop over last dim.

Public Attributes inherited from GpuLaunchConfig
cudaFuncAttributes	attr
	attributes of the function

Detailed Description

3D launch configuration for symmetry-reduced G-space loops (z dimension folded for real data sets)

The documentation for this struct was generated from the following file:

Public Member Functions