Recently, the National Science Foundation (NSF) and NVIDIA jointly announced an investment of $152 million to launch a new project called "Open Multimodal AI Infrastructure for Accelerating Science" (OMAI), led by the Allen Institute for Artificial Intelligence (AI2). The project aims to develop open-source multimodal AI models to help researchers in various fields conduct scientific research more efficiently.
In the OMAI plan, the National Science Foundation will provide $75 million, while NVIDIA will invest $77 million to create an open AI ecosystem. AI2 will focus on training domain-specific multimodal large language models based on scientific literature. These models will help researchers accelerate their research, generate code and visual content, and connect new discoveries with previous research findings.
Through the OMAI models, researchers can design new materials faster, assist in climate modeling, or help biologists discover protein interactions, all of which require integrating text, data, and image information. NVIDIA will also provide HGX B300 systems equipped with Blackwell Ultra GPUs and an AI enterprise software platform to support the smooth progress of this project.
The collaborating institutions include the University of Washington, the University of Hawaii at Hilo, the University of New Hampshire, and the University of New Mexico. The participation of these research institutions ensures that cutting-edge AI technology is not limited to a few top laboratories, but can benefit a broader range of researchers.
Unlike many proprietary models, OMAI promises to openly release models, training data, code, evaluation methods, and documentation, which will help improve the transparency and reproducibility of scientific research. Additionally, the project aligns with the White House's AI Action Plan, emphasizing the importance of open science in enhancing U.S. competitiveness and addressing concerns about AI bias and accountability.
For enterprise IT leaders, the OMAI project marks a shift in the way AI infrastructure is built and shared, promoting a transition from closed ecosystems to open collaboration models. The project may influence enterprises' choices between cloud services and on-premises infrastructure, compliance and governance, as well as the demand for skills in maintaining open models.
Although the investment in OMAI is relatively small, it represents a shared foundation of public resources that can promote collaboration among universities, startups, and companies. The implementation of this project will provide research institutions, especially laboratories with limited funding, with more advanced AI tools, thus accelerating innovation.
Key Points:
🌟1. The National Science Foundation of the United States and NVIDIA have invested $152 million to jointly develop open AI models, improving the efficiency of scientific research.
🔍2. OMAI will create an open-source AI ecosystem, promoting collaboration and discovery among researchers in multiple fields.
🚀3. The project will help lower the barriers to scientific research, promote the popularization of AI technology, and ensure that more institutions can access cutting-edge technological support.
