New York Advances Research Computing Infrastructure
New York state has officially activated its $40 million Empire AI Beta supercomputer. This system, located at the State University of New York at Buffalo, provides academic researchers across a wide consortium of institutions with significantly increased computing power for data-intensive projects. More than 300 research initiatives are already waiting to access the infrastructure, which marks the second major phase of the $500 million Empire AI project.
The hardware package represents a substantial upgrade from the initial Alpha system released in 2024. It features four NVIDIA GB200 NVL72 systems, which includes 288 NVIDIA Blackwell GPUs, alongside 60 NVIDIA Grace CPU Superchip nodes. The facility also hosts 40 petabytes of storage capacity. Consortium leaders report that this setup delivers 11 times the AI training capacity and 40 times the inference capacity of the previous machine.
A Collaborative Approach to Academic Computation
The Empire AI initiative operates on a shared resource model rather than anchoring the technology within a single university. Participants include major entities such as the City University of New York, Columbia University, Cornell University, and New York University, alongside the Icahn School of Medicine at Mount Sinai and the Flatiron Institute. This structure allows smaller or specialized institutions to pursue projects that were previously impossible due to prohibitive costs and physical infrastructure requirements.
Juan J. de Pablo, Executive Vice President for Global Science and Technology at New York University, noted that access to high-powered computing previously restricted the questions researchers could reasonably address. With this new capacity, scholars can now look into medical diagnostics, climate modeling, and cybersecurity with higher precision. The project team believes this shift moves the state closer to its goals of creating faster discovery cycles in advanced manufacturing and other technical fields.
Expansion Plans and Long-term Sustainability
The current Beta system is an interim step toward a more permanent solution. Construction is already underway for the final Gamma facility at the University at Buffalo, which is slated for completion by the end of 2027. This next-generation building will feature a closed-loop thermal energy system designed to repurpose waste heat from the computing equipment to warm other campus facilities, aligning the project with broader net-zero environmental targets.
State officials are integrating this effort with the National Science Foundation's State and Regional AI Infrastructure Hubs program. By aligning local academic goals with federal infrastructure strategies, New York seeks to stabilize the costs of sustaining such high-performance hardware. While 300 projects wait in the current queue, the anticipated completion of the Gamma site in 2027 will further determine how many researchers the state can support at once. The focus remains on building an environment where computational limits do not prevent academic breakthroughs.