Joseph Cooke, research computing team member, won Northeastern’s 2025 Staff Excellence Award for “Innovation Across the Network.”
After spending eight years in graduate school and earning a Ph.D. in mechanical engineering and applied mechanics, Joseph Cooke realized he enjoyed supporting other scientists more than pursuing his own academic research.
This discovery led Cooke to a job in research computing at Northeastern University, where he’s worked for nearly three years.
This year, Cooke, a computational scientist in the research computing department in the Office of the Provost, received the highest number of nominations for a 2025 Staff Excellence Award at Northeastern in the “Innovation Across the Network” category.
The awards selection team noted that nomination letters portrayed Cooke as a transformative force, quietly revolutionizing operations and making a significant impact in his role in a small department.
“His influence and the ripple effect of his efforts are remarkable and inspiring,” the selection team wrote.
Northeastern’s research computing team manages the high-performance computing cluster located in the Massachusetts Green High Performance Computing Center in Holyoke, Massachusetts. The cluster is available to all Northeastern faculty and students to use at no cost.
An HPC cluster is a powerful system made up of many interconnected computers working together to solve computationally intensive problems that are far beyond what a single computer can handle. The research computing team supports Northeastern researchers across many disciplines, from artificial intelligence and machine learning to quantum physics, material science or psychology.
“We welcome any project that needs computational power,” Cooke says. “We work with researchers, whether they are postdocs, professors, grad students or undergrads and try to help make sure their work appropriately fits onto the cluster.”
Cooke’s role is to ensure researchers have the right tools, hardware and software, on the cluster. He also helps troubleshoot and improve research workflows.
“We do get some of the more theoretical questions where they might not be as familiar with programming [compared to their own area of expertise],” Cooke says.That’s why the research computing team includes domain experts like him — scientists with experience in big data, bioinformatics, data analytics, machine learning or molecular dynamics — who can understand research projects and make meaningful contributions. Cooke himself has a background in computational chemistry, physics and engineering.
The team works remotely, with members spread across states like California, Arizona, and Michigan, supporting faculty and students across Northeastern’s global network — from Oakland to North Carolina to London. Cooke is based in his home state of Pennsylvania, near Philadelphia.
One of Cooke’s favorite parts of the job is solving unusual user requests, especially when he needs to determine whether an issue is specific to the application or related to the system itself and requires installation of new software.
“We all like helping people,” he says about his team. “We are excited when we are actually able to enable different kinds of research on the cluster.”
The team collaborates on a lot of projects and holds frequent meetings throughout the day on such topics as improving documentation, making changes to the website or discussing possible new policies to keep the HPC system running smoothly.
The HPC cluster can be used not only for research but also to support educational courses. However, when Cooke joined Northeastern, the process of setting up new course environments and adding students to the cluster was complex and time-consuming.
He’s proud of helping automate that process by connecting the cluster to Canvas, Northeastern’s learning management system. Now, students enrolled in a course can be identified automatically and have accounts created on the cluster.
“That is one of the really big projects because it not only helps us, but it helps professors get access for their students in a much faster way,” he says. “We’re always happy to meet with faculty to see if using the cluster is a good fit for their class.”
Another effort Cooke is proud of was a solo project done with a graduate co-op student. Cooke decided to test if there was a performance difference between running different science applications on the cluster servers versus packaging, or containerizing, them so they can run on a regular local computer.
“You’re not going to have the same capability as the HPC, but for testing and debugging, it’s fantastic,” Cooke says. “If we containerize software, the researchers are able to take their research anywhere and connect to other clusters they have access to.”
Over four months, Cooke and his assistant ran tests to measure how well different scientific applications performed under certain conditions — how fast they ran and how much memory they used. Their results improved containerization documentation and provided the team with clear data they can confidently share with researchers about what to expect when using containers.
“Without that project, we might have gotten those types of requests and it would have taken longer to resolve,” Cooke says. “I’m proud of it because I led the effort from start to finish and created something that truly benefited the team.”
Cooke emphasizes that his team is always open to feedback from clients because it helps them continuously improve their services.
When he is off the clock, Cooke likes to step away from the computer screens and spend time hiking with his dog, attending car racing events, and traveling along the East Coast and to national parks.