[ACCEPTED]-Compiling CUDA with clang-llvm-clang

Accepted answer

Score: 11

Thanks to contributions from Google and 7 others, Clang now supports building CUDA. Command 6 line parameters are slightly different from 5 nvcc, though. According to the official documentation, assuming your 4 file is named axpy.cu, the basic usage is:

$ clang++ axpy.cu -o axpy --cuda-gpu-arch=<GPU arch>  \
    -L<CUDA install path>/<lib64 or lib>              \
    -lcudart_static -ldl -lrt -pthread

Note 3 that using Clang for compiling CUDA still 2 requires that you have the proprietary CUDA 1 runtime from the NVIDIA CUDA toolkit installed.

Score: 4

2016-05-01 Update: clang now supports CUDA. See @rivanvx' answer.

The CUDA compiler is based on LLVM. Clang, though 1 also based on LLVM, does not support CUDA.

Source: stackoverflow.com

More Related questions

how to add include paths to clang globally

What do you need to install to use Clang on windows to build c++14 for 64 bit?

Can C++17 be used together with CUDA using clang?

Are llvm-gcc and clang binary compatible with gcc? - particularly mingw gcc on Windows

Get member of __m128 by index?

GCC(/Clang): Merging functions with identical instructions (COMDAT folding)

Can parameter pack function arguments be defaulted?

Clang 14 and 15 apparently optimizing away code that compiles as expected under Clang 13, ICC, GCC, MSVC

How to print dependencies from llvm / clang (equivalent of gcc -MD)?

Clang OS X Lion, cannot find cstdint

Cookie Warning

We use cookies to improve the performance of the site. By staying on our site, you agree to the terms of use of cookies.