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
+12 -3
View File
@@ -33,7 +33,7 @@ target_link_libraries(
PUBLIC
CUTLASS
cutlass_tools_util_includes
$<$<BOOL:${CUTLASS_ENABLE_CUBLAS}>:cublas>
$<$<BOOL:${CUTLASS_ENABLE_CUBLAS}>:nvidia::cublas>
gtest
)
@@ -48,6 +48,8 @@ target_link_libraries(
PUBLIC
cutlass_test_unit_infra
)
set(CUTLASS_INSTALL_TESTS ON CACHE BOOL "Install test executables")
function(cutlass_test_unit_add_executable)
@@ -65,7 +67,7 @@ function(cutlass_test_unit_add_executable)
PRIVATE
cutlass_test_unit_infra
cutlass_test_unit_infra_lib
)
)
string(REGEX REPLACE cutlass_ "" NAME_STEM ${NAME})
@@ -79,7 +81,14 @@ function(cutlass_test_unit_add_executable)
${NAME}
)
# message(STATUS "cutlass_test_unit_add_executable(${NAME} c${NAME_STEM} ${NAME_STEM})")
if (CUTLASS_INSTALL_TESTS)
install(
TARGETS ${NAME}
RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR}
)
endif()
endfunction()