Nicholas Nystrom, PhD
Chief Technology Officer
Nicholas A. Nystrom, PhD, is Chief Technology Officer at Peptilogics. In this role he leads technology planning, implementation and operation for the following areas: artificial intelligence (AI), machine learning (ML), and computational science, strategic computing and data systems including future technologies. He also leads technology partnerships and is active in university relations.
Prior to joining Peptilogics, Nick was Chief Scientist at the Pittsburgh Supercomputing Center (PSC), a joint research center of Carnegie Mellon University and the University of Pittsburgh. While at PSC, Nick developed a complete, sustainable high-performance computing (HPC), AI and data ecosystem supporting research in medicine, science and engineering, in coordination with diverse federal sponsors (NSF, NIH, DoD) and partnerships with academia and industry. He architected and was principal investigator (PI) and project director for supercomputers that served the national and international research community, examples of which are Blacklight, the world’s largest shared-memory computer; Bridges, which in 2014 pioneered the convergence of AI, HPC and big data and was the world’s first computer designed to use the Intel Omni-Path Interconnect, deployed in a custom topology to maximize performance for data-intensive science; and Bridges-2, which scaled technologies proven in Bridges and added an all-flash filesystem, hierarchical data management and quadrupled bandwidth for deep learning, optimized to accelerate rapidly-evolving science. He also co-architected Bridges-AI, which increased the aggregate national AI capacity of NSF cyberinfrastructure by 283%, and Neocortex, a unique architecture comprised of a large-memory (24TB) front-end optimized for and tightly coupled to two Cerebras CS-1 servers, designed to enable research on scaling AI beyond a single wafer-scale engine.
Specific to the life sciences, Nick was PI for the Human BioMolecular Atlas Program (HuBMAP) program (NIH), creating data and computational hardware and software infrastructure for developing a map of all tissues of the human body at single-cell resolution, with data spanning genomic, proteomic and imaging modalities. He led early work to establish interoperability between HuBMAP and other medical data initiatives involving childhood cancers, structural birth defects and nerve-organ interactions. Nick has also led work in machine learning for breast and lung cancer (Pennsylvania) and causal discovery focusing on cancer driver mutations, lung fibrosis and the brain causome (an NIH Big Data to Knowledge Center of Excellence). In 2020, Nick was closely involved in the White House COVID-19 High Performance Computing Consortium, through which Bridges was made available to support urgent research on the emerging virus.
In other fields, Nick has led R&D in computational chemistry and advanced computing (DoD), computer architecture (DoD/DOE OSC/DOE NNSA), computer languages (DARPA/IBM), ab initio molecular dynamics (NSF), performance engineering (NSF), combustion kinetics (DOE), high energy nuclear physics (DOE), nuclear engineering (Westinghouse) and hardware/software performance (NASA/NSF).
Nick has organized international conferences focusing on AI, HPC and computational science, and he has served on and chaired numerous federal review panels. Nick has presented numerous invited talks on advanced computing, AI and computational science, and he has published in areas of computer architecture, deep learning and machine learning applications, genomics, data architecture, ontology design, performance engineering and high-productivity programming languages.
Nick earned his PhD in Chemistry (specializing in quantum chemistry) at the University of Pittsburgh, also completing graduate coursework in Physics. Prior to that, he majored in Chemistry, Math and Physics.