Enum cublasGemmAlgo_t
An enumerant to specify the algorithm for matrix-matrix multiplication Complete documentation on Nvidia documentation
Namespace: Hybridizer.Runtime.CUDAImports
Assembly: Hybridizer.Runtime.CUDAImports.dll
Syntax
[IntrinsicType("cublasGemmAlgo_t")]
public enum cublasGemmAlgo_t
Fields
Name | Description |
---|---|
CUBLAS_GEMM_ALGO0 = 1 | Explicitly choose Algorithm 0 |
CUBLAS_GEMM_ALGO0_TENSOR_OP = 20 | Explicitly choose GEMM Algorithm 0 while allowing the use of Tensor Core operations when possible |
CUBLAS_GEMM_ALGO1 = 2 | Explicitly choose Algorithm 1 |
CUBLAS_GEMM_ALGO1_TENSOR_OP = 21 | Explicitly choose GEMM Algorithm 1 while allowing the use of Tensor Core operations when possible |
CUBLAS_GEMM_ALGO10 = 11 | Explicitly choose Algorithm 10 |
CUBLAS_GEMM_ALGO11 = 12 | Explicitly choose Algorithm 11 |
CUBLAS_GEMM_ALGO12 = 13 | Explicitly choose Algorithm 12 |
CUBLAS_GEMM_ALGO13 = 14 | Explicitly choose Algorithm 13 |
CUBLAS_GEMM_ALGO14 = 15 | Explicitly choose Algorithm 14 |
CUBLAS_GEMM_ALGO15 = 16 | Explicitly choose Algorithm 15 |
CUBLAS_GEMM_ALGO16 = 17 | Explicitly choose Algorithm 16 |
CUBLAS_GEMM_ALGO17 = 18 | Explicitly choose Algorithm 17 |
CUBLAS_GEMM_ALGO2 = 3 | Explicitly choose Algorithm 2 |
CUBLAS_GEMM_ALGO2_TENSOR_OP = 22 | Explicitly choose GEMM Algorithm 2 while allowing the use of Tensor Core operations when possible |
CUBLAS_GEMM_ALGO3 = 4 | Explicitly choose Algorithm 3 |
CUBLAS_GEMM_ALGO4 = 5 | Explicitly choose Algorithm 4 |
CUBLAS_GEMM_ALGO5 = 6 | Explicitly choose Algorithm 5 |
CUBLAS_GEMM_ALGO6 = 7 | Explicitly choose Algorithm 6 |
CUBLAS_GEMM_ALGO7 = 8 | Explicitly choose Algorithm 7 |
CUBLAS_GEMM_ALGO8 = 9 | Explicitly choose Algorithm 8 |
CUBLAS_GEMM_ALGO9 = 10 | Explicitly choose Algorithm 9 |
CUBLAS_GEMM_DEFAULT = 0 | Apply Heuristics to select the GEMM algorithm |
CUBLAS_GEMM_DEFAULT_TENSOR_OP = 19 | Apply Heuristics to select the GEMM algorithm, and allow the use of Tensor Core operations when possible |