Georgia Tech is set to unveil its Nexus supercomputer, a groundbreaking system designed to revolutionize artificial intelligence (AI) and high-performance computing (HPC). Backed by a $20 million investment from the National Science Foundation (NSF), the university has committed to launching the system in spring 2026. The Nexus, with a capacity of 400 petaflops—equivalent to solving 50 million math problems for every person on Earth every second—will be tailored to AI research and scientific challenges.
A key feature of the Nexus is its extensive memory and storage, offering 330 terabytes of RAM and 10 petabytes of flash storage, equivalent to the digital storage of 10 billion reams of paper. This robust infrastructure is critical for training large AI models, running complex simulations, and managing large datasets. The system is designed with a balanced approach, ensuring efficiency and usability, with built-in user-friendly interfaces to support scientists without deep technical expertise.
With Nexus, Georgia Tech is not only advancing its own research but also enabling broader national access through a hybrid model that allocates 10% of the system for on-campus use while the NSF manages national access. This dual-access approach is expected to foster innovation across various fields, including health, energy, climate science, and more. The impact of the Nexus extends beyond academia, with potential applications in drug discovery, vaccine development, improved weather predictions, and smarter energy systems.
By opening up access to advanced AI infrastructure, Nexus aims to democratize scientific research, enabling a wider range of researchers to conduct experiments and collaborate across disciplines. This move by Georgia Tech is a significant step in making cutting-edge technology more accessible, thereby accelerating discoveries that could address global challenges in the near future.