Skip to content

NSF OMAI: Ai2 puts in the air AI infrastructure fully open for the academy

Manu Ramalho
Manu Ramalho

The Allen Institute for AI (Ai2) has activated the NSF OMAI computing infrastructure, providing open AI resources to the academic community.

Loading audio…
Cover image for: NSF OMAI: Ai2 puts in the air AI infrastructure fully open for the academy

The Allen Institute for AI (Ai2) has just put online the computing infrastructure NSF OMAI, making a fully open AI research ecosystem a reality. Now laboratories and universities in the United States – and by extension, the world – can access high-performance computational power without traditional barriers of cost and bureaucracy, accelerating discoveries in artificial intelligence.

What is the NSF OMAI and how does this open AI infrastructure differ?

OMAI (Open Multi-Access Infrastructure) is an initiative of the National Science Foundation (NSF) in partnership with Ai2. Unlike commercial cloud services, it has been designed from the beginning with the "open by design" principle: all hardware, software and workflows are open and replicable. This means that researchers not only use computational power, but can also inspect, modify and share infrastructure components.

The infrastructure includes state-of-the-art GPU clusters, high-performance storage and experiment orchestration tools. The goal is to democratize access to resources that were once unique to large technology companies such as Google, Meta and OpenAI.

How do researchers access this open AI infrastructure?

Any researcher affiliated with a U.S. academic institution may request access to the OMAI through a simple project proposal process. Unlike previous initiatives, there is no need for commercial partnerships or approval of complex committees. Once approved, the researcher obtains an account with computer credits, similar to a cloud model, but without direct costs.

Transparency is a differential: the entire cluster configuration is documented and published in open repositories, allowing other groups to replicate the environment locally. This strengthens scientific reproducibility, one of the biggest challenges of AI research today.

Why is this open AI infrastructure important to Brazil?

Although initial access is focused on American institutions, the open by design model serves as a global reference. Countries like Brazil, which seek sovereign AI infrastructure, can be inspired by OMAI to create their own open clusters. In addition, international collaboration projects often allow Brazilian researchers to participate through partnerships with US universities.

The open AI movement is also reflected on other fronts, such as the launch of Meta Llama 3.1 405B, the largest open model to date, which redefines the boundaries of open-source. Together, these advances signal an irreversible trend: AI research is becoming less dependent on infrastructure monopolies.

Frequently asked questions about the open AI infrastructure NSF OMAI

Who can request access to the NSF OMAI?

Researchers affiliated with academic or non-profit research institutions in the United States may submit project proposals. The process is open and peer reviewed.

Is OMAI really free?

Yes, for approved projects, the use of infrastructure has no direct cost. The NSF covers operational expenses as part of the investment in scientific research.

How does OMAI compare to services like AWS or Google Cloud?

While commercial services charge per hour GPU and have closed service level agreements, OMAI is fully open and cost-free for the academy, focusing on reproducibility and transparency.

The Ai2 NSF OMAI initiative represents a concrete step towards a more inclusive and collaborative AI science. With open infrastructure, the future of research no longer depends on billion-dollar budgets – it depends on good ideas.

For more information on the subject, check out our complete article on /news/nsf-omai-ai2-infrastructure-open.

Share:

Source: allenai.org

Manu Ramalho

Written by

Manu Ramalho

Sou Manu Ramalho, publicitária com 16 anos de estrada conectando marcas e pessoas. Como fundadora da EME Marketing Digital, sempre busquei o marketing estratégico para gerar conexões autênticas. Aqui, mergulho na fronteira da inteligência artificial como analista de tendências. Meu foco é traduzir a complexidade de NLP, novos modelos de linguagem e papers acadêmicos para o mundo real, sempre com um olhar atento à regulamentação, ética e aos impactos sociais que essa tecnologia imprime na nossa sociedade.

Related articles