OLCF and NCCS Director

Director’s Letter

As an applied scientist with three decades of AI experience in biomedical discovery and healthcare delivery, I have seen first-hand the importance of high-performance computing. This personal experience is why I am so proud to lead the National Center for Computational Sciences (NCCS), a division of the Computing and Computational Sciences Directorate at ORNL and home to the Oak Ridge Leadership Computing Facility (OLCF). NCCS and the OLCF provide an unrivaled collaborative environment to push the boundaries of computational science, especially as we blaze a trail toward new realms of high-performance computing power.

As both a division within the Oak Ridge National Laboratory and a US Department of Energy (DOE) Leadership Computing Facility, our organization’s mission is to provide world-class computational resources and specialized services for researchers at ORNL and around the globe. OLCF users come from diverse scientific domains such as materials science, earth sciences, astrophysics, and biology, where they tackle the most computationally complex problems. In addition, the OLCF responds as needed at times of national and global crisis. The COVID-19 pandemic is a perfect example of how the OLCF lends its world-leading resources and cutting-edge computing expertise to accelerate understanding of this disease and its possible treatments. For example, the OLCF is a founding member of the COVID-19 HPC Consortium supporting multiple large-scale R&D efforts on bioinformatics, drug discovery, and epidemiology across academia, industry, and federal agencies.

In executing our mission, we also recognize the dynamic nature of computational science and will always readily adapt to its changing needs. We are currently experiencing a tipping point with scientific advances happening at a rapid pace. AI has been transforming the ways we think about and conduct science. Scientific methodologies are advancing and that will inevitably influence the nature of the OLCF as a user facility, because our user community will evolve in terms of how they leverage emerging computing infrastructures.

While simulations provided the bulk of computing workloads in years past, the computational problems we’re tackling today are exceedingly data intensive. For more than a decade, the scientific community has been dealing with a tsunami of data-generating technologies. Data-driven scientific discovery has taken an accelerated turn and the trend will continue with growing advances in artificial intelligence. NCCS and the OLCF are committed to pioneering secure HPC systems for data-intensive workloads, just as we have spearheaded computational advances for compute-intensive modeling and simulation workloads.

Even as we prepare our facilities for the 2021 launch of what will become the world’s fastest computer, Frontier, and its projected peak performance of 1.5 exaflops, we know that computational problems will grow ever larger in the 21st century. We have a wealth of experience in building predictive models by using supercomputers to solve complex first-principles physics equations. With the steadily increasing power of today’s supercomputers and the massive data sets that are becoming available in a variety of areas, we are now in a position to build the same types of predictive models using AI.

Our goal is to not only effectively support both communities, but also to enable advances that bridge these two worlds. Large-scale data analysis interleaved with large-scale simulations is an opportunity that lies ahead for our division, for our leadership computing center, for Frontier and beyond.

Georgia Tourassi

Director, National Center for Computational Sciences