Georgia Institute of Technology, along with its partners, has received a $20 million grant from the National Science Foundation (NSF) to build Nexus, an advanced supercomputer designed to accelerate breakthroughs in areas like medicine, clean energy, climate science, and robotics.
Set to go live in spring 2026, Nexus will be one of the fastest AI-powered supercomputers in the country. It will be capable of performing more than 400 quadrillion operations per second. To put this into perspective, it’s the equivalent of every person on Earth doing 50 million calculations each second without stopping.
Georgia Tech’s president Ángel Cabrera said, “Georgia Tech is proud to be one of the nation’s leading sources of AI talent and technologies that are powering a revolution in our economy. It’s fitting that they selected us to host this new supercomputer, and it will support a new wave of AI-centered innovation across the nation. We’re grateful to the NSF, and we are excited to get to work.”
Supercomputers that are often only accessible to major research hubs, Nexus will be open to researchers from universities and institutions across the country. Georgia Tech will oversee access to the system through an NSF-managed application process, ensuring that even smaller institutions have a chance to work with one of the most powerful AI systems in the world. Georgia Tech will reserve ten percent of Nexus’ capacity for its own campus-based research projects.
Nexus is designed to handle some of the toughest scientific challenges of our time. From finding new drugs to fighting climate change and building smarter robots. The system’s enormous computing power and advanced storage capacity will speed up discoveries.
Katie Antypas, director of the NSF Office of Advanced Cyberinfrastructure, explained the system’s unique capabilities. “The Nexus system’s novel approach combines support for persistent scientific services with more traditional high-performance computing. It will enable new science and AI workflows that will accelerate the time to scientific discovery,” she said. “We look forward to adding Nexus to NSF’s portfolio of advanced computing capabilities for the research community.”
Nexus will house an enormous amount of memory and storage. In particular, the system will include 330 terabytes of memory and 10 petabytes of flash storage. That’s enough digital space to hold the same amount of data as 10 billion reams of paper. If people stacked that paper, it would reach about 500,000 kilometers. That distance would stretch from Earth to the Moon and partway back.
Nexus is not just about size. It will use ultra-fast data transfer systems. This allows researchers to move large datasets quickly between systems. It reduces waiting time significantly.
This speed is essential for complex simulations and massive data projects. For instance, these include climate modeling and human genetics research.
In this context, Suresh Marru, principal investigator of the Nexus project and director of Georgia Tech’s Center for AI in Science and Engineering (ARTISAN), emphasized its importance for researchers worldwide. Specifically, “This supercomputer will help level the playing field,” he said. Moreover, “It’s designed to make powerful AI tools easier to use and available to more researchers in more places.”
Nexus is part of a broader national effort to strengthen America’s leadership in artificial intelligence and supercomputing. Georgia Tech is building the system in partnership with the National Center for Supercomputing Applications (NCSA) at the University of Illinois Urbana-Champaign. When Nexus is completed, it will connect to NCSA’s systems through a new high-speed network. This will create shared infrastructure that helps scientists across the U.S. collaborate more effectively.
Charles Isbell, chancellor of the University of Illinois and former dean of Georgia Tech’s College of Computing, called it a symbol of teamwork. He said Nexus is more than a supercomputer. It represents what is possible when leading institutions work together to advance science.
This project builds on Georgia Tech’s earlier work with its precursor system, HIVE, and the CODA data center that will house Nexus. Earlier in 2024, Georgia Tech partnered with Nvidia to deploy an AI supercomputer hub for students. It features 20 Nvidia HGX H100 systems and 160 Nvidia H100 Tensor Core GPUs.
Srinivas Aluru, Regents’ Professor and senior associate dean in the College of Computing, described Nexus as the result of years of careful planning. In fact, he said Nexus places Georgia Tech among academic supercomputing centers. He added that it reflects years of planning, including the CODA data center and the earlier HIVE project.
