For decades, high performance computing (HPC) has been the primary driver of advanced system architecture and platform innovation; today, AI workloads are increasingly pushing the bleeding edge of technology. In this session, we look beyond strategy to examine how these shifts are translating into real server and HPC solutions, and consider how the knock on impact of “RAMageddon” could force a rethink of traditional approaches to system sizing and solution design.