NVIDIA NvNeural SDK
2022.2
GPU inference framework for NVIDIA Nsight Deep Learning Designer
|
Helper macros and ILibraryContext functions for working with cuBLAS. More...
Go to the source code of this file.
Macros | |
#define | CheckCublasErrors(expr_) |
CheckCublasErrors is an error-handling macro. More... | |
Functions | |
cublasHandle_t | nvneural::GetCublasHandle (const INetworkBackendCuda *pBackend) noexcept |
Retrieves the cuBLAS handle associated with a CUDA backend. More... | |
NeuralResult | nvneural::InitializeCublas (INetworkBackendCuda *pBackend) noexcept |
Initializes cuBLAS and registers it with the backend. More... | |
Helper macros and ILibraryContext functions for working with cuBLAS.
#define CheckCublasErrors | ( | expr_ | ) |
CheckCublasErrors is an error-handling macro.
It returns NeuralResult::Failure from the caller if the provided cublasStatus_t expression evaluates to anything but CUBLAS_STATUS_SUCCESS.
Example: CheckCublasErrors(cublasSgemm(...))
|
noexcept |
Retrieves the cuBLAS handle associated with a CUDA backend.
This uses the ILibraryContext system.
pBackend | CUDA backend to query |
|
noexcept |
Initializes cuBLAS and registers it with the backend.
This uses the ILibraryContext system.
pBackend | CUDA backend to enable for cuBLAS |