United States: AI Capabilities and Prospects

Traditional HPC architectures have typically been designed to address either homogenous workloads (such as physics-based modeling and simulation) with similar, and perhaps more important, fixed, compute, memory, and I/O requirements or, more recently, heterogenous workloads with a diverse range of compute, memory, and I/O requirements. Most HPC data center planners and operators, however, don't have the luxury of focusing on one main type of workload; they typically must support many HPC users and their associated workloads sporting a wide range of compute, memory, and I/O profiles. Architectures have typically consisted of a fixed set of resources, resulting in an underutilized system with expensive elements often sitting idle for long periods of time. One approach being explored to increase system utilization by exposing resources that would otherwise sit idle to appropriately matched jobs waiting in a queue is via composable systems.

Surpassing pre-pandemic attendance and achieving new attendance and exhibitor participation records, SC23 in Denver, CO did not disappoint. With over 14,000 on-site attendees and over 435 exhibitors, SC23 provided a vibrant, high-energy environment for the global HPC community to exchange ideas while establishing or extending collaborative partnerships and relationships. As is our custom, the Hyperion Research team of analysts has compiled its primary takeaways and perspectives from the event.