In this insideHPC Guide, our friends over at WEKA suggest that when organizations decide to move existing applications or new applications to a GPU-influenced system there are many items to consider, such as assessing the new  environment’s required components, implementing a pilot program to learn about the system’s future  performance, and considering eventual scaling to production levels.