You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
On the LANL venado system, cuda installation is provided by the NVIDIA HPC SDK. The cuda toolkit libraries are located at /opt/nvidia/hpc_sdk/Linux_aarch64/24.7/cuda, while the cuda math libraries are located separately at /opt/nvidia/hpc_sdk/Linux_aarch64/24.7/math_libs.
In spack, cuFFT is not defined as a separate package but is instead searched at the location of the cuda toolkit. However, on venado, since the cufft (and other math libraries) are not co-located with the cuda toolkit, any application (e.g. lammps) that depends on cuFFT fails to build as it cannot find the required include and lib files.
The text was updated successfully, but these errors were encountered:
On the LANL venado system, cuda installation is provided by the NVIDIA HPC SDK. The cuda toolkit libraries are located at
/opt/nvidia/hpc_sdk/Linux_aarch64/24.7/cuda
, while the cuda math libraries are located separately at/opt/nvidia/hpc_sdk/Linux_aarch64/24.7/math_libs
.In spack,
cuFFT
is not defined as a separate package but is instead searched at the location of thecuda
toolkit. However, on venado, since the cufft (and other math libraries) are not co-located with the cuda toolkit, any application (e.g. lammps) that depends on cuFFT fails to build as it cannot find the required include and lib files.The text was updated successfully, but these errors were encountered: