Georgia Tech building $20M AI supercomputer that outcomputes 8 billion humans

Scheduled for completion by spring 2026, Nexus will be among the most powerful AI research supercomputers in the US, delivering 400 quadrillion operations per second and enabling nationwide access for breakthroughs in healthcare, energy, climate, neuroscience, and robotics

In a landmark move aimed at accelerating artificial intelligence (AI)-powered research, the United States National Science Foundation (NSF) has granted $20 million (approximately ₹166 crore) to the Georgia Institute of Technology to develop a cutting-edge supercomputer named Nexus. The initiative underscores America’s growing focus on AI-led scientific innovation, and reflects a global trend of investing in AI infrastructure to address pressing challenges across industries.

Once completed by spring 2026, Nexus is expected to become one of the most powerful AI research systems in the US, capable of performing over 400 quadrillion operations per second. Housed at Georgia Tech’s advanced CODA data center, it will provide open access to scientists across the US, allowing them to tackle grand challenges such as drug discovery, climate change modelling, clean energy solutions, neuroscience, robotics, and quantum materials.

Nexus sets global AI benchmark

Georgia Tech President Ángel Cabrera said the system will usher in a new era of AI-driven research. “Nexus will be a transformative tool, and we are proud to be at the forefront of this movement,” he stated.

For Indian academics, scientists, and students observing international developments, Nexus represents a significant benchmark in combining AI and supercomputing for national-scale research. India too has been expanding its AI infrastructure under initiatives like the National Supercomputing Mission (NSM), but the scale and AI specificity of Nexus highlight where global standards are headed.

The system will feature 330 terabytes of memory and 10 petabytes of flash storage, enabling seamless handling of large scientific datasets. A collaborative effort with the National Center for Supercomputing Applications (NCSA) at the University of Illinois Urbana-Champaign will ensure high-speed connectivity between institutions, forming a national research backbone.

Democratising access to AI power

What sets Nexus apart is its design: it merges traditional high-performance computing with persistent AI services, enabling scientists to run complex models efficiently. Access will be granted through an NSF-reviewed application process, democratizing high-end AI capabilities across institutions.

Suresh Marru, the project's principal investigator and director of Georgia Tech’s new Center for AI in Science and Engineering, noted, “This supercomputer will make powerful AI tools more accessible, especially to researchers from less-resourced institutions.”