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 -5
View File
@@ -22,16 +22,14 @@
set(CUTLASS_EXAMPLES_COMMON_SOURCE_DIR ${CMAKE_CURRENT_SOURCE_DIR}/common)
function(cutlass_example_add_executable)
function(cutlass_example_add_executable NAME)
set(options)
set(oneValueArgs)
set(multiValueArgs)
cmake_parse_arguments(_ "${options}" "${oneValueArgs}" "${multiValueArgs}" ${ARGN})
cutlass_add_executable(${__UNPARSED_ARGUMENTS})
list(GET __UNPARSED_ARGUMENTS 0 NAME)
cutlass_add_executable(${NAME} ${__UNPARSED_ARGUMENTS})
target_link_libraries(
${NAME}
@@ -59,7 +57,10 @@ foreach(EXAMPLE
05_batched_gemm
06_splitK_gemm
07_volta_tensorop_gemm
08_turing_tensorop_gemm)
08_turing_tensorop_gemm
10_planar_complex
11_planar_complex_array
)
add_subdirectory(${EXAMPLE})
add_dependencies(cutlass_examples ${EXAMPLE})