ROOT 6.16/01 Reference Guide |
The TDevice class provides static functions for the generation of CUDA grids for kernel launches and is used to encapsulate the distribution of threads and blocks over the data.
Static Public Member Functions | |
static dim3 | BlockDims1D () |
static dim3 | BlockDims2D () |
template<typename AFloat > | |
static dim3 | GridDims1D (const TCudaMatrix< AFloat > &A) |
template<typename AFloat > | |
static dim3 | GridDims2D (const TCudaMatrix< AFloat > &A) |
template<typename AFloat > | |
static int | NThreads (const TCudaMatrix< AFloat > &A) |
Static Public Attributes | |
static constexpr int | BlockDimX = 1 |
static constexpr int | BlockDimY = 32 |
static constexpr int | BlockSize = BlockDimX * BlockDimY |
#include <TMVA/DNN/Architectures/Cuda/Device.h>
|
inlinestatic |
|
inlinestatic |
|
inlinestatic |
|
inlinestatic |
|
inlinestatic |
|
staticconstexpr |
|
staticconstexpr |