Internet2 Deployed New Cisco AI Infrastructure
The research network added modular AI compute nodes to two university campuses to bolster national processing power.
Updated on Sept. 22, 2026 in Data Centers

Live Poll
Do you believe shared national computing infrastructure improves the quality of research in your area?
Internet2 has installed new Cisco AI PODs at the University of Nebraska-Lincoln and Sacramento State to expand the National Research Platform. These modular units add significant GPU and CPU resources to a shared fabric used by researchers across the United States.
Why it matters
This standardized infrastructure lowers barriers for researchers by simplifying the complex engineering required to integrate new computing resources. By aggregating capacity, the platform enables broader access to advanced AI tools for students and faculty.
The deployment provides 1.1 TB of combined GPU memory, 2 TB of system memory, and 90 TB of NVMe flash storage. Each AI POD supports connectivity speeds of up to 200 Gbps to ensure high-performance data throughput.
The players
Internet2
This non-profit organization operates a high-speed research network serving the academic and research community.
Cisco
This multinational technology corporation provides networking hardware and modular AI infrastructure solutions.
The details
The National Research Platform utilizes a Kubernetes-based fabric called Nautilus to manage resources contributed by more than 70 organizations. The new PODs provide pre-validated hardware, allowing institutions to seamlessly integrate high-performance computing into the existing nationwide network.
Timeline
Internet2 was founded in 1996.
The expansion of the National Research Platform was announced on September 22, 2026.
The Tech Race
The adoption of standardized, pre-validated AI PODs signals a shift toward modularity in research computing, replacing fragmented custom-built environments. This strategy mirrors the broader industry move to streamline high-performance computing clusters through software-defined orchestration.
Researchers and students at participating institutions gain faster, more standardized access to high-end GPU resources for AI-driven projects. This reduces wait times for computing jobs and lowers the technical overhead for teams managing data-intensive workflows.
The takeaway
Standardizing research infrastructure is essential as the demand for AI compute capacity continues to rise across higher education. By pooling hardware resources, institutions can effectively bridge the digital divide for specialized academic research.
Further reading
Learn more about the latest innovations in Data Centers to understand how research networks manage increasing computational demands.
More information
For additional information, visit the Internet2 official organizational website.
Live Poll
Do you believe shared national computing infrastructure improves the quality of research in your area?






