As discussed at the recent CMS meeting. This is part 2 of possible multi-GPU development. Support for multi-GPU nodes (2: send MEs to more than one GPU from the same madevent executable?) Note: - part 1 in #836 is simply about choosing a single GPU (should be easy, and useful) - part 2 in #989 is about sending MEs to more than one GPU from the same executable (complex, and not necessarily useful?) - part 3 in #990 is about orchestrating from python/bash the different madevent executables to target different GPUs (not to ocomplex, and sounds very useful?)