Cheaha is more than just a supercomputer, it’s a growth opportunity. UAB harnesses the power to help the research community grow — and it’s part of what drew Carlos Cardenas, Ph.D, to become a Blazer himself.
Cardenas arrived at UAB in June 2021, working in the Heersink School of Medicine’s Department of Radiation Oncology. He is currently the director of automated treatment planning, an assistant professor, and serves as director of AI research and development in the Marnix E. Heersink Institute for Biomedical Innovation. Before joining UAB, Cardenas worked at the University of Texas M.D. Anderson Center. He served on the advisory group for high-performance computing.
At UAB, he’s worked closely with UAB IT’s research computing team, led by Ralph Zottola, Ph.D., assistant vice president for research computing.
“One of the first phone calls I made when I got the job here was to Ralph,” Cardenas said. “He was instrumental in confirming my decision to come to UAB. I felt good about the department I was joining, and he assured me the computing resources I needed would be there when I arrived.”
One of the first things Cardenas did when arriving on campus was create his account in Cheaha. His lab was also one of the firsts to test out the earliest version of cloud.rc. At the time, the lab’s jobs were bigger than what the cloud could handle. When the newer queues became available to use, the lab began utilizing the high-performance computing cluster more.
Cardenas’s lab works on several projects with Research Computing, with his trainees using the cluster the most. Currently, the lab is working on radiotherapy treatment plans and focusing on dosage predictions for patients.
“We start the process once a patient has joined our clinic,” Cardenas said. “They get scans, and from there we meet with their physician, symmetrist, and physicists like me. We discuss where the radiation needs to come from and what needs to be treated versus not treated. From start to finish this could take a few weeks to complete.”
Cardenas’ team and collaborators have a strong background in software development and have been able to develop a system to cut down the time from days to hours. Treatment plans can be assessed and handed back to an oncologist within 30 minutes.
“Anytime we come up with a clever idea or need to test something before integration with a clinical workflow, we try to run simulation in Cheaha,” Cardenas said. “A lot of our success has been in the supercomputer. It’s nice to know we can send over 100 jobs and get answers in a matter of minutes.”
If there is a moment of uncertainty, all Cardenas’s team has to do is jump on a quick call with the Research Computing team. Every week, UAB IT Research Computing holds virtual office hours for researchers to pop in and ask questions. Cardenas urges his trainees and anyone else wanting to start using Cheaha to jump on a call and talk to the UAB IT scientists.
“My students take advantage of these services tremendously,” Cardenas said. “If they come to me with a problem I always ask if they’ve talked to Research Computing first. We’ve had a very positive experience in communicating and collaborating with the department.”
Another thing Cardenas praises the department for is resources. Along with long-term storage, and an evolving list of upgrades, Research Computing also makes it easy for anyone to access the cluster. Cardenas is “old school” and finds it easiest to log in via SSH, while students lean more towards accessing Cheaha in a web application. Resources like this, even if it just a small matter, is one way that the department supports the university, and UAB’s Growth with Purpose Initiative.