Aurora, to be hosted by Argonne National Laboratory, is one of three planned exascale-class systems in the U.S. While the Intel-led system has encountered a variety of conceptual transformations (it was originally planned as a pre-exascale system) and setbacks (Intel’s delayed 7nm node), that hasn’t stopped excited researchers from planning their use of the massive system. Among those researchers is a team led by Amanda Randles, an assistant professor of biomedical sciences at Duke University, that is preparing to utilize Aurora to visualize how cancer spreads through the human body.

The tool in question is called HARVEY, which has previously been used to track how microscopic blood cells flow through our bodies. The researchers are repurposing those same capabilities to also model microscopic cancer cells, hoping to use the newfound power of Aurora to both create a more accurate model of the circulatory system and understand how tumor cells move through that model. “By understanding the biological mechanisms behind metastasized cancer cells, too, we hope our work with HARVEY will eventually help doctors and their patients in the fight against cancer,” Randles said in an interview with Argonne’s Joan Koka.

The team has plenty of time before Aurora arrives, but also plenty of hurdles to clear before that point: for instance, how to account for the large disparity between Aurora’s computational speeds and its write-to-disk speeds. The researchers, like many others facing the same problem, are looking to perform at least some of the data analysis within the machine (rather than exporting and analyzing the data) to help ameliorate the issue.

“If, for example, you identify a particular area of interest within the whole system, you might choose to save data at a higher frequency or higher level of detail just in that small area. That would still enable you to get more science out of the data while reducing the amount of data that you actually have to write to disk,” said Joseph Insley, lead of the visualization and data analysis team at the Argonne Leadership Computing Facility (ALCF). 

“Tackling our new research into the process of metastasis and performing the intricate simulations needed means we need even greater computing power to handle the massive data sets in real time. The Aurora system will help us meet this need,” added Randles.

The ALCF has also been preparing a unified library of frameworks for this on-machine data analysis and visualization process (part of a DOE project called SENSEI), which ALCF staff are working to ready for the HARVEY code base using the ALCF’s existing supercomputer power.

“We’ve been using the ALCF’s Theta supercomputer and early Aurora hardware to integrate this library with the HARVEY code,” Insley said. “By enabling visualization and analysis on the data while it’s still in memory, the science team will be able to gain more insights from the data than they otherwise would.”

Randles’ team is one of 15 selected for the ALCF’s Aurora Early Science Program (ESP). Aurora is scheduled for delivery in 2022.

To read the reporting from Argonne’s Joan Koka, click here.