High-performance computing for research is notorious for having steep barriers to entry. For this reason, high-tech disciplines were early adopters and heaviest users of HPC resources. But as more fields engage with data-intensive research and artificial intelligence (AI) workflows, performance technologies are evolving to accommodate a larger and more diverse community of practice.
A new resource known as the Interactive Data Analytics Service (IDAS) is making advanced computational capabilities more accessible to University of Iowa researchers. The free toolkit is especially appealing to those who haven’t used HPC before, but whose work benefits from the power of a supercomputer.
When data-analytics applications are accessed through the IDAS interface, they look and feel just like they do on a regular workstation, but researchers gain access to significantly more computing power. IDAS has its own HPC and graphics processing units (GPUs) and allows researchers to perform interactive data-analysis tasks with applications used for machine learning and AI.
“The IDAS platform provides access to processing power that isn’t otherwise available to many of us who aren’t traditional HPC users,” says Strength and Conditioning Coach Landon Evans of UI Intercollegiate Athletics and Sports Science, whose research involves the forecasting of athlete performance, injury, and other risk outcomes.
Since local computing resources in Athletics are needed for other tasks, Evans appreciates being able to off-load the more compute-intensive tasks to IDAS, without having to purchase commercial cloud computing cycles. He says he’s looking forward to exploring the GPU capabilities of IDAS in the future.
ITS Research Services launched IDAS in August 2019. Since then, the number of users and classes registered to use this novel platform has doubled.
IDAS supports large-scale and collaborative data analytics workflows using RStudio for R and Jupyter Notebook for Python, R, and Julia. IDAS’ Jupyter Notebook serves both general research and classroom needs, which were optimized during its pilot phase. A dozen faculty and staff from six departments test-drove the research platform for several months, and classroom attributes were trialed by students in a summer business-analytics course taught by Data Scientist Kang Lee of ITS Research Services and the Tippie College of Business.
“I use IDAS’ Jupyter Notebooks with Python 3 for my research on social media group dynamics,” says Computational Sociologist Yongren Shi of the UI College of Liberal Arts and Sciences. “While I haven’t used it in the classroom yet, I recommended it to graduate students in our department. I appreciate that it offers flexibility in the number of compute nodes I can create myself, and that it’s free.”
In early January, a Data Science Institute co-sponsored by the Iowa Social Science Research Center (ISRC) and ITS-Research Services introduced participants to the basic building blocks of coding as they’re applied in a variety of research domains, including social sciences and the humanities. Nearly 250 students, faculty and staff attended the trainings, and dozens created accounts to start using IDAS.
At first Associate Professor of Biology Sarit Smolikove was hesitant to attend the institute, for fear her programming skills weren’t up to speed, but co-organizer Sai Kumar Ramadugu of ITS-Research Services put her fears to rest.
“IDAS demystified the act of engaging with HPC,” Smolikove says. “The trainers were helpful, and I was glad I attended.”
Next steps for IDAS
In the future, IDAS developers will assist with custom environments and a remote desktop feature will be added. There are also plans to facilitate access to Iowa’s Argon supercomputer if more power is needed and to purchase commercial cloud computing power if the demand is there.