New York has switched on a new academic supercomputer designed to give researchers across the state access to artificial intelligence hardware previously available mainly to technology companies and elite research laboratories. Governor Kathy Hochul announced that Empire AI Beta is now fully operational at the State University of New York at Buffalo, marking a major expansion of the state’s public-interest artificial intelligence program. The $40 million, NVIDIA-powered system is intended to accelerate research in medicine, climate science, cybersecurity, education, advanced manufacturing and public safety while keeping the development of AI focused on socially beneficial applications.
Empire AI Beta is the second generation of the state’s shared AI computing infrastructure. Compared with Empire AI Alpha, the initial system introduced in 2024, Beta provides an 11-fold increase in AI training capacity, a 40-fold improvement in AI inference performance and an eight-fold expansion in data storage. Training is the computationally intensive process through which machine-learning models analyze enormous datasets and adjust billions of internal parameters. Inference occurs after training, when a model generates predictions, diagnoses or other outputs. Increasing both capabilities allows researchers to build larger models and use them more rapidly in real-world experiments.
More than 300 research projects are already queued to use Empire AI Beta. The system will allow scientists to run simulations, analyze complex biological data and develop machine-learning tools without relying exclusively on commercial cloud providers or limited university computing clusters. In health research, for example, high-performance AI systems can compare medical images, identify patterns in genomic data and assist in the design of new treatments. Climate scientists can use them to model extreme weather and changing ecosystems, while engineers can apply them to infrastructure monitoring, advanced materials and safer transportation systems.
The scale of the new system is significant because AI research is increasingly constrained by computing power. Modern models require vast quantities of data and repeated calculations across specialized graphics processing units, or GPUs. These processors can perform many mathematical operations simultaneously, making them well suited to neural networks, which represent information through layers of interconnected numerical weights. By pooling resources through Empire AI, researchers from public universities, private institutions and nonprofit laboratories can undertake projects that would be too expensive or technically demanding for a single institution to support independently.
“New York State built Empire AI to show that artificial intelligence can be developed for the public good,” Hochul said in announcing the launch. She described Beta as the most powerful academic AI research computer in the country and said the project demonstrated how government, universities, philanthropy and industry could cooperate to build advanced scientific infrastructure. Empire AI Board Chairman Tom Secunda similarly characterized the system as a shared scientific asset, emphasizing that its purpose is to advance research rather than serve the commercial interests of one organization.
The new computer is also attracting attention beyond New York. The National Science Foundation has announced a State and Regional AI Infrastructure Hubs initiative intended to expand shared computing capacity and support regional research networks. New York officials say Empire AI provides an operating model for that national effort by combining public investment, academic expertise and philanthropic support. Rather than concentrating AI resources in a small number of technology companies, the approach distributes access among universities and researchers working on projects with potential public benefits.
Empire AI Alpha has already supported more than 130 research projects involving hundreds of researchers across New York. Beta represents the next stage of a larger buildout, while construction continues on the permanent Gamma facility at the University at Buffalo. Gamma is expected to be completed by the end of 2027 and is planned as an even larger and more efficient high-performance computing center. Its design will connect the computing facility to the university’s thermal energy network, allowing heat generated by the processors to be captured and used to warm campus buildings. This closed-loop arrangement could reduce wasted energy and help the university meet its net-zero targets.
The consortium behind Empire AI includes the State University of New York, the City University of New York, Columbia University, Cornell University, New York University, Rensselaer Polytechnic Institute, the University of Rochester, the Rochester Institute of Technology, the Icahn School of Medicine at Mount Sinai and the Flatiron Institute at the Simons Foundation. The project is supported by more than $500 million in public and private funding. Expanded participation is intended to give researchers at institutions across the state access to advanced computing, including scientists working in regions that may not otherwise be able to finance specialized AI infrastructure.
University leaders said the system could broaden opportunities for students and faculty as well as established research groups. SUNY Chancellor John B. King Jr. said Empire AI would support work in fields such as health care, public safety and emerging technologies, while CUNY Chancellor Félix V. Matos Rodríguez described the initiative as a way to place advanced computing in the hands of a more diverse academic community. Research Computing Director Kiran Keshav said the system would enable scientists to ask larger questions, test more ambitious ideas and move faster from theoretical work to practical discoveries. With Beta now online, New York is positioning shared, publicly oriented AI infrastructure as a central component of its research strategy.
Subject of Research: Shared artificial intelligence infrastructure and high-performance computing for academic research and public-interest applications.
Article Title: New York Activates Empire AI Beta, Opening Advanced Supercomputing to Researchers Across the State
Web References: https://www.nsf.gov/news/new-nsf-state-regional-ai-infrastructure-hubs-will-power-ai ; https://www.governor.ny.gov/news/governor-hochul-signs-legislation-expanding-historic-empire-ai-consortium-and-establishing ; https://www.governor.ny.gov/news/governor-hochul-launches-nations-first-independent-ai-research-center-public-university-state ; https://www.suny.edu/
References: Office of Governor Kathy Hochul; State University of New York; Empire AI consortium.
Image Credits: State University of New York
Keywords
Empire AI, artificial intelligence, supercomputing, NVIDIA, machine learning, academic research, New York, SUNY Buffalo, AI infrastructure, climate science, medical AI, public-interest technology
Tags: advanced AI model training and inferenceAI applications in cybersecurity and educationAI hardware accessibility for researchersAI in medicine and climate scienceArtificial intelligence supercomputing infrastructureequitable access to AI technologyexpansion of AI data storage capacitygovernment-supported AI initiativeshigh-performance computing for AI researchNVIDIA-powered AI systemspublic-interest AI research programsstate-funded AI acceleration projects
