Classic 2022

The design space for large scale distributed computing and analytics is increasingly rich, complex, and diverse (e.g. Accelerator technologies, IoT, SmartNICs, Computational Storage, Hybrid-MultiCloud, and HPC).

As scale and distribution grows, so does the impact of where data is moved and computations occur, as well as the need to address resilience, control, privacy, and security. The build out of the Internet of Things (IoT) and other large-scale, geographically dispersed federated systems (such as smart cities, connected transportation, power grids, and healthcare) presents a multitude of opportunities and challenges when designing large-scale analytics. Optimally managing up to tens of billions of endpoints capable of collecting new and growing sources of data, and millions of heterogeneous distributed computing elements (e.g., sensors, edge processors, mobile devices, on-premise servers, SmartNICs, Computational Storage, Clouds, HPC systems, etc.) creates significant performance, cost, and software development complexities. Insightful and pivotal decisions on where to collect, process, and analyze data will be required to achieve acceptable latencies, accurate conclusions, at acceptable costs.