diff options
| author | Patrick Mours <pmours@nvidia.com> | 2020-10-13 13:33:34 +0300 |
|---|---|---|
| committer | Brecht Van Lommel <brecht@blender.org> | 2021-02-08 17:50:03 +0300 |
| commit | c661515090b189061c4388a335573ac3d5745925 (patch) | |
| tree | 5f81489cb4906b5d3fa0252c37c897cb3bc36383 /CMakeLists.txt | |
| parent | 8d02177ab079eec518685d60f5e0d9e146de97cb (diff) | |
Cycles: Add CUDA 11 build support
With this patch the build system checks whether the "CUDA10_NVCC_EXECUTABLE" CMake
variable is set and if so will use that to build sm_30 kernels. Similarily for sm_8x kernels it
checks "CUDA11_NVCC_EXECUTABLE". All other kernels are built using the default CUDA
toolkit. This makes it possible to use either the CUDA 10 or CUDA 11 toolkit by default and
only selectively use the other for the kernels where its a hard requirement.
Reviewed By: brecht
Differential Revision: https://developer.blender.org/D9179
Diffstat (limited to 'CMakeLists.txt')
| -rw-r--r-- | CMakeLists.txt | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/CMakeLists.txt b/CMakeLists.txt index 6f705ffbe44..b15bbb7486b 100644 --- a/CMakeLists.txt +++ b/CMakeLists.txt @@ -369,7 +369,7 @@ option(WITH_CYCLES_CUDA_BINARIES "Build Cycles CUDA binaries" OFF) option(WITH_CYCLES_CUBIN_COMPILER "Build cubins with nvrtc based compiler instead of nvcc" OFF) option(WITH_CYCLES_CUDA_BUILD_SERIAL "Build cubins one after another (useful on machines with limited RAM)" OFF) mark_as_advanced(WITH_CYCLES_CUDA_BUILD_SERIAL) -set(CYCLES_CUDA_BINARIES_ARCH sm_30 sm_35 sm_37 sm_50 sm_52 sm_60 sm_61 sm_70 sm_75 compute_75 CACHE STRING "CUDA architectures to build binaries for") +set(CYCLES_CUDA_BINARIES_ARCH sm_30 sm_35 sm_37 sm_50 sm_52 sm_60 sm_61 sm_70 sm_75 sm_86 compute_75 CACHE STRING "CUDA architectures to build binaries for") mark_as_advanced(CYCLES_CUDA_BINARIES_ARCH) unset(PLATFORM_DEFAULT) option(WITH_CYCLES_LOGGING "Build Cycles with logging support" ON) |