In the 3DEXPERIENCE platform the GPU based calculation of global illumination is sometimes not working properly on multi-GPU workstation.
This can happen if two different GPU architectures are used within the same system (e.g. NVIDIA RTX GPU and NVIDIA Pascal GPU).
To solve the problem the user needs to specify which GPU should be used for CUDA calculations. This can be done by setting the environment variable CUDA_VISIBLE_DEVICES to the proper GPU identifier.
An official specification for this variable can be found within the CUDA programming guide from NVIDIA: https://docs.nvidia.com/cuda/cuda-c-programming-guide/index.html#env-vars