In California, a new system has emerged that could revolutionize US biological defense strategies. The Pentagon, in collaboration with the National Nuclear Security Administration, has developed a cutting-edge complex that integrates a powerful supercomputer with a state-of-the-art, rapid response laboratory (RRL).
The cornerstone of this system is a supercomputer based on the groundbreaking El Capitan architecture – the future exaflops behemoth of the Livermore National Laboratory (LLNL). This supercomputer is powered by advanced AMD Mi300A APU accelerators, which combine high-performance CPUs and GPUs.
The primary objective of this new complex is to enhance both military and civilian defense against biological threats. It will conduct large-scale simulations, develop drugs, and utilize AI tools for modeling and categorizing potential hazards.
It is worth mentioning that not only the military but also other government agencies of the United States, international allies, academic institutions, and industry will benefit from this platform. This broad access is essential due to the wide range of potential dangers that could impact civilian populations, water resources, food supplies, and other critical areas.
The Rapid Response Laboratory, located in close proximity to the computing center, complements the capabilities of the GenieTrained Intelligent Drug Engineering program, designed to develop novel medical treatments. This program employs machine learning to create antibodies, analyze experimental data, explore structural biology, and conduct molecular simulations.
Thanks to this new system, the Department of Defense and three national laboratories (LLNL, Sandia, and Los Alamos) will be able to accelerate the vaccine and antibody development process. Scientists can now swiftly test numerous drug options that were initially designed through computer simulations.
The RRL laboratory is equipped with cutting-edge technology, including automated robots capable of altering protein structures and producing prototypes of new medications simultaneously. Researchers can identify vulnerabilities in disease pathogen structures much more efficiently.
“We are not focusing on a single project or specific biological or chemical threat. We are establishing a national infrastructure to swiftly assess and respond to any potential dangers, irrespective of the nature of future threats,” said Jim Breis, LLNL’s Deputy Director for Computational Technology.