A new National Science Foundation grant is helping the University of Miami expand research networking and AI capacity at smaller institutions across the region along with a partnership with Florida LambdaRail.

Regional Cyberinfrastructure

As artificial intelligence reshapes higher education and the workforce, the University of Miami is emerging as a regional hub for the computing power, infrastructure, and training that many smaller South Florida universities cannot yet provide on their own.

Now, with a $1.4 million grant from the National Science Foundation, Ravi Vadapalli, PhD, director of advanced computing at the Frost Institute for Data Science and Computing (IDSC), is leading an effort to help roughly two dozen smaller universities across the region build the capacity to keep pace.

The two-year grant will strengthen regional advanced-networking infrastructure, expand research capacity, and support Dr. Vadapalli’s broader vision for collaboration among South Florida institutions.

“My vision is to build a true regional consortium for AI, data science, and research computing,” he said.

For Dr. Vadapalli, the challenge is not just that some institutions have fewer resources. It is also that students across the region are preparing for the same workforce without equal access to the tools, training environments, and research opportunities that increasingly define education in AI and data science.

“If we want the region to grow,
we need our neighboring institutions
to do well too.”

“At UM, we have strong facilities, computing resources, and the ability to train students with state-of-the-art tools,” Dr. Vadapalli said. “Students at institutions without those resources enter the same workforce, but they have not had equal access to the same training environment. If we want the region to grow, we need our neighboring institutions to do well too.”

A Partnership Fosters Connection

Many smaller universities lack the network capacity needed to connect to high-performance computing resources. Others struggle to match their infrastructure with research needs, estimate computing costs for teaching and research, or build competitive proposals around cyberinfrastructure.

Through a planning grant, Dr. Vadapalli created a series of workshops last year for faculty and students at institutions with limited computing backgrounds. Those sessions helped participants identify research and teaching needs, think through proposal development, and better understand how data science and AI could fit into their campuses’ missions.

The work also laid the groundwork for broader collaboration with Florida LambdaRail, Florida’s research and education network, and other partner institutions, building on conversations that continued through regional planning meetings last year.

What became clear through those sessions was that internet access alone was not enough. Many institutions still lacked the infrastructure needed to support advanced research computing. “We discovered that many institutions did not have the network capacity to connect to the computing services we can provide at UM,” he said.

“We discovered that many institutions
did not have the network capacity
to connect to the computing services
we can provide at UM.”

The grant is designed to address that gap by helping partner institutions connect to Florida LambdaRail’s 400 Gigabits-per-second research network, which supports the large-scale data transfer, collaboration, and scientific computing that commercial internet service often cannot. The University of Miami is one of 13 higher education institutions that own Florida LambdaRail, making the partnership a natural extension of existing research infrastructure in the state.

“It brings these smaller, under-resourced institutions into the fold so they can participate in research, collaborate beyond their campuses to advance science and technology, and provide high-quality research opportunities for their students” said Jon Ellis, CEO of Florida LambdaRail and the grant’s co-principal investigator.

This connectivity will shape not only whether programs in areas like bioinformatics, imaging, and sequencing analysis are feasible, but also what happens in the classroom. “When I taught a class at Florida Memorial back in 2022, I had to prerecord lectures because live demonstrations weren’t feasible given their network limitations,” Dr. Vadapalli said.

Phase One

In the first phase of the project, partner institutions like St. Thomas University and Florida Memorial University will receive support to upgrade their networking so they can connect to Florida LambdaRail and access higher-capacity research infrastructure.

Through the grant, those campuses will receive 10-Gigabits-per-second connections to the network, significantly expanding the bandwidth available for data-intensive research instruction. In addition, Bethune-Cookman University is upgrading its legacy networking infrastructure through the grant. The University of Miami will provide technical support, as it already does for its own faculty and students, so the schools can use the resources effectively.

Dr. Vadapalli sees the consortium
as a pipeline for future collaborations,
grant proposals, and external funding
that can benefit the entire region.

Dr. Vadapalli sees the consortium as a pipeline for future collaborations, grant proposals, and external funding that can benefit the entire region. Institutions that may not have been competitive on their own can become stronger through shared planning, infrastructure, and support. In one early example, multiple institutions came together around a collaborative proposal worth $1.6 million after conversations that grew out of this work. “That’s exactly the kind of derivative benefit this consortium can create,” he said.

Sustainability is also central to the plan. Connecting campuses to the research network is only the first step. Keeping them connected will require long-term institutional commitment and continued funding.

“Many institutions may offer courses in data science, but without adequate infrastructure and faculty support, they may not be able to provide students with the most current or effective training,” he said. “That affects the overall return on public investment in higher education.”

Dr. Vadapalli is working to help campus leaders understand that long-term value while also identifying diverse funding opportunities to sustain the infrastructure.

A Model for the Future

The project also introduces a “Science DMZ in a box” model designed to make advanced networking more manageable for smaller campuses with limited IT staffing and resources. A Science DMZ is a network architecture built for secure, high-performance scientific data transfer. Under this model, partner campuses maintain simpler on-site equipment, while Florida LambdaRail manages the more complex connectivity and external components through the collaboration. The goal is to give institutions the benefits of advanced connectivity without asking them to maintain systems they realistically cannot support on their own.

“If, in the future, NSF creates an opportunity
for a major AI center or cyberinfrastructure center
in South Florida, I want us to be ready.”

At the national level, Dr. Vadapalli said, the National Science Foundation is making major investments in AI workforce development and research infrastructure. His hope is that South Florida will be prepared when larger opportunities emerge.

“If, in the future, NSF creates an opportunity for a major AI center or cyberinfrastructure center in South Florida, I want us to be ready,” Dr. Vadapalli said. ◙

 

Story by Lauren Comander