The National Institute of Health’s premier supercomputer, Biowulf, is the largest supercomputer in the world dedicated to biomedical research and life sciences. Today, the agency is using it to address the planet’s most urgent public health challenge: How do we beat coronavirus?
GDIT began upgrading the Biowulf system four years ago, transforming the system to be ranked #67 top supercomputer in the world in 2017. Our modernization and expansion efforts include the addition of 100,000 compute cores - delivering more than two thousand trillion operations per second - and 35 petabytes of online storage. In 2018 alone, Biowulf ran more than 34 million jobs and delivered more than 1 billion CPU hours for the scientific research community.
Today, Biowulf is used by more than 3,000 researchers – including nearly half of the agency’s principal investigators – and has been cited in over 2,500 research publications and supports more than 600 scientific applications and tools.
GDIT’s successful architecture, design and deployment of four expansion phases delivered a powerhouse system ready to meet the demands of today’s pandemic research.
Since early March 2020, NIH researchers studying COVID-19 have used the Biowulf supercomputer to examine the structure of the virus, study its genetic variations and make predictions about community spread, dedicating 28.5 million CPU hours to process over 834,000 jobs (as of November 2020).
NIH researchers have also been able to perform a wide range of computational studies including modeling of unreported infections, investigation of pathogenic characteristics, RNA structural analysis, transmission modeling, and identification of drugs that can be repurposed for COVID-19 treatments.
The computational scale of Biowulf allows NIH scientists to ask questions on a much larger scale and on more complicated problems than historically possible – because Biowulf enables NIH to support multiple, complex research endeavors at once and accelerate the speed at which the research is conducted.
The supercomputer is designed to process a large number of simultaneous computations, those common in fields like genomics, image processing, statistical analysis and other biomedical research areas. This research can lead to new treatments for cancer, diabetes, heart disease, mental health and infectious diseases including COVID-19.
When NIH made the significant, multi-year investment in high-performance computing, they could not have predicted the global pandemic that would hit just a few years later. That investment is proving to be a sound one, expanding the agency’s computational capabilities and ability to scale computational analysis quickly and efficiently when it is needed the most.