CHARMM Development Project Forums User Discussion & Questions Installation and Testing Installation of c48 on ubuntu 22.04

Hello CHARMM users and developers,

I am facing an issue while building CHARMM on my Ubuntu 22.04 due to the gcc version. Specifically, I am using gcc 9.5.0 along with cuda 11.0, and encountering an error.

nvcc warning : The 'compute_35', 'compute_37', 'compute_50', 'sm_35', 'sm_37' and 'sm_50' architectures are deprecated, and may be removed in a future release (Use -Wno-deprecated-gpu-targets to suppress warning).
In file included from /apps/cuda/cuda-11.0/include/cuda_runtime.h:83,
from :
/apps/cuda/cuda-11.0/include/crt/host_config.h:139:2: error: #error -- unsupported GNU version! gcc versions later than 9 are not supported! The nvcc flag '-allow-unsupported-compiler' can be used to override this version check; however, using an unsupported host compiler may cause compilation failure or incorrect run time execution. Use at your own risk.
139 | #error -- unsupported GNU version! gcc versions later than 9 are not supported! The nvcc flag '-allow-unsupported-compiler' can be used to override this version check; however, using an unsupported host compiler may cause compilation failure or incorrect run time execution. Use at your own risk.

I would like to know if there is a way to install CHARMM with this current setup or if I should upgrade or downgrade my gcc and cuda versions. Could you kindly suggest which versions are advisable to use? I would greatly appreciate any help and suggestions you can provide.

Thank you.

I don't know because the System Requirements section of the document 'Installation Guide Linux' for cuda toolkit 11.0 says it supports GCC 9.x
https://docs.nvidia.com/cuda/archiv...ide-linux/index.html#system-requirements

I think this is a bug in a header file provided by the cuda toolkit.

If it were me, I would do what the error message says and try GCC version 8.

The default gcc on Ubuntu 22 is version 11, so another possibility is the environment was not set up correctly, either for the 9.5 compiler, or for CUDA.

After the compiler environment is set up, what is the result of the command 'gcc --version' ?