Hank Childs is an academic expert in high performance computing and scientific visualization. At the University of Oregon, he is an associate professor of computer and information science. Hank’s primary research focus in on visualizing and analyzing the very large data sets generated by physics ...
Hank Childs is an academic expert in high performance computing and scientific visualization. At the University of Oregon, he is an associate professor of computer and information science. Hank’s primary research focus in on visualizing and analyzing the very large data sets generated by physics simulations on supercomputers. This task has become increasingly challenging, since supercomputers’ computational ability has recently outpaced its storage ability; as a result visualization is being performed “in situ” and visualization programs must make a decision about how to proceed with no human in the loop. Further, supercomputer architectures are shifting rapidly, now with thousands of cores per node and deep memory hierarchies, among other changes, requiring research to re-think how visualization algorithms can run effectively in this environment.