High performance computing, and advanced computing in general, face a plethora of opportunities and challenges. The opportunities include: new hardware options (new CPUs, new competitors in the GPU space, and hardware being created and assembled as part of various exascale initiatives); new software particularly focused on containers, cloud computing, and AI; and new opportunities in applications such as biomedical research and climate studies. There are a plethora of new challenges however, including the fact that FAANG (Facebook, Apple, Amazon, Nvidia, Google) are collectively outspending the US government on CS R&D; security and privacy challenges; and a public that is ever more skeptical of basic aspects of the scientific method and the outcomes of federally funded scientific research.
This talk will focus on national challenges and opportunities in four areas of HPC-oriented R&D topics relevant to NIST’s historic mission related to computing, and outline particular needs and challenges where NIST leadership would be highly valuable to the nation: biomedical research and security of personal information (with a focus on cancer research); AI; exascale parallel computing; and reproducibility of software analyses. This talk will hopefully be interesting in surveying certain areas of accomplishment (biomedical research is after all the reason the speaker is alive to be giving this talk) and also suggest opportunities that may be interesting to NIST’s experts as topics for future NIST activities.