Cuda build runs out of memory on default runner #1817
Description
Comment:
When trying to enable Cuda support in conda-forge/exchcxx-feedstock#2 (upstream wavefunction91/ExchCXX#46), the CI runner is going out of memory when compiling the Cuda code.
Does anyone know how this can be solved with the existing CI runners or whether it is necessary to have a runner with larger memory for this kind of builds?