High-Performance Statistical Computing: Current Trends and Future Prospects

• Activity Number: 24 - High-Performance Statistical Computing: Current Trends and Future Prospects
• Type: Topic Contributed
• Date/Time: Sunday, August 7, 2022 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistical Computing
• Author(s): Marc Genton* and Dorit Hammerling* and Zhengqing Ouyang* and George Ostrouchov* and Hatem Ltaief*
• Companies: KAUST and Colorado School of Mines and University of Massachusetts, Amherst and ORNL and KAUST
• Keywords: HPC; HPSC; supercomputer; big data; large-scale; statistical computing

Abstract

High-Performance Computing (HPC) is having a tremendous impact on scientific applications from different fields such as chemistry, medicine, engineering, and space research. HPC can help solve complex problems that were previously deemed impossible to solve. Recently, HPC has been involved in tackling big data challenges in statistical applications, and many examples can be provided to demonstrate the power of HPC in handling and analyzing statistically big datasets. However, the statistics community is still far from exploiting all the advantages of existing HPC tools and hardware. More efforts are needed to improve the community's awareness about HPC benefits and how it can open the door to solve many challenges in existing scientific applications. We intend in this panel to discuss the current efforts to use HPC capabilities to solve complex problems in different statistics branches, setting a clear definition for the High-Performance Statistical Computing (HPSC) term. Moreover, the panel will discuss the community expectations about HPSC and how it can shape the future of statistical computing.