CUTLASS 2.1 (#83)

CUTLASS 2.1 contributes:
- BLAS-style host-side API added to CUTLASS Library
- Planar Complex GEMM kernels targeting Volta and Turing Tensor Cores
- Minor enhancements and bug fixes
This commit is contained in:
Andrew Kerr
2020-04-07 13:51:25 -07:00
committed by GitHub
parent 7c0cd26d13
commit 96dab34ad9
196 changed files with 20653 additions and 1995 deletions
+6 -3
View File
@@ -23,9 +23,13 @@
cutlass_test_unit_add_executable(
cutlass_test_unit_gemm_device
BATCH_SOURCES ON
BATCH_SIZE 4
gemm_f16t_f16n_f16t_tensor_op_f16_sm75.cu
gemm_f16n_f16t_f16t_tensor_op_f16_sm75.cu
gemm_f16n_f16t_f16t_tensor_op_f16_sm75_slicedk.cu
gemm_f16n_f16t_f16t_tensor_op_f16_slicedk_sm75.cu
gemm_f16t_f16n_f16t_tensor_op_f16_slicedk_sm75.cu
gemm_f16n_f16n_f16t_tensor_op_f32_sm75.cu
@@ -90,6 +94,7 @@ cutlass_test_unit_add_executable(
simt_zgemm_tn_sm50.cu
simt_zgemm_tt_sm50.cu
gemm_splitk_serial_tensor_op_sm75.cu
gemm_splitk_tensor_op_sm75.cu
gemm_splitk_tensor_op_sm70.cu
gemm_splitk_simt_sm50.cu
@@ -145,5 +150,3 @@ cutlass_test_unit_add_executable(
gemm_f16t_f16n_f32t_singlestage_wmma_tensor_op_f32_sm70.cu
)