NVIDIA partners with the U.S. National Science Foundation to create an AI system supporting the development of multimodal language models for scientific research in the U.S. The partnership, part of the NSF Mid-Scale Research Infrastructure project, aims to accelerate scientific breakthroughs and secure U.S. global leadership in science and technology.

OMAI, a project of the Allen Institute for AI, aims to build a national open AI ecosystem to drive scientific discovery through AI. NVIDIA’s support includes providing state-of-the-art AI infrastructure and software platform to transform datasets into actionable intelligence and breakthrough innovations. The collaboration aims to empower U.S. scientists with cutting-edge infrastructure for research.

NVIDIA HGX B300 systems, featuring industry-leading technologies, will provide groundbreaking acceleration, scalability, and efficiency for running large models and demanding workloads. The investment supports research teams from various universities and aligns with the White House AI Action Plan to support America’s global AI leadership.

Large language models (LLMs) with billions of parameters are driving applications, but access to these models can be limited for scientific research without openly available data, code, and documentation. NVIDIA’s partnership with NSF and Ai2’s OMAI initiative aims to provide open model access, data tools, and training for researchers to advance U.S. global leadership in science and engineering.

The OMAI initiative aligns with the White House AI Action Plan to accelerate AI-enabled science and promote leading open models to enhance America’s global AI leadership. The project aims to support academic research and education by providing access to open-source software and models, in line with Ai2’s commitment to maximizing access to cutting-edge technology.

Read more at NVIDIA: NVIDIA, National Science Foundation Support Ai2 Development of Open AI Models to Drive US Scientific Leadership