What just happened? The US National Science Foundation and Nvidia have partnered to build advanced artificial intelligence infrastructure designed to support scientific research across the country. Backed by significant funding and cutting-edge AI technology, the collaboration aims to help scientists solve complex research challenges and maintain the United States' leadership in global AI innovation.
A central element of the partnership is funding the Open Multimodal AI Infrastructure to Accelerate Science (OMAI) project. The NSF will contribute $75 million, while Nvidia is providing $77 million in resources and technology.
The project is led by the Allen Institute for AI (AI2) and will leverage large volumes of scientific data to build advanced AI models. These models aim to make it easier for researchers to analyze information, generate code, create visualizations, and uncover insights across fields such as materials science, biology, and energy.
The OMAI initiative is a key component of NSF's Mid-Scale Research Infrastructure program, which supports critical infrastructure projects that fall outside the scope of standard federal grants or major national laboratories.
Beyond building new AI tools, the program also focuses on workforce development, training researchers from institutions outside major tech hubs. This approach is designed to broaden access to advanced AI and strengthen the US workforce in high-impact technology areas.
Nvidia will equip OMAI with cutting-edge AI computing platforms, including Nvidia HGX B300 systems powered by Blackwell Ultra GPUs. These GPUs deliver ultra-fast processing, high-speed memory, and enhanced interconnectivity – capabilities essential for training and deploying frontier multimodal language models, which demand enormous computational resources. The project will also leverage Nvidia AI Enterprise software to simplify deployment, management, and scaling of AI tools for research.
These advanced multimodal models can process not only text, but also images, graphs, and tables, making them highly effective for complex scientific applications. However, ensuring open access to these models, along with their training data, code, and documentation, remains critical to broad adoption across academic and research communities.
The collaboration also aligns with the White House AI Action Plan, issued in July, which calls for increased federal support for AI-driven science, accelerated data center permitting, and global leadership in core AI technologies. The NSF-Nvidia partnership addresses these priorities by providing the compute power and infrastructure needed to advance open AI models for scientific research.
OMAI will support research teams at multiple institutions, including the University of Washington, the University of Hawaii at Hilo, the University of New Hampshire, and the University of New Mexico.
By democratizing access to high-performance AI models and infrastructure, the initiative aims to accelerate breakthroughs in areas such as biomedical prediction, materials science, and energy research.
Nvidia partners with the National Science Foundation to accelerate AI projects
