In an interview, ACM Fellow and ACM/IEEE Kennedy and Athena award recipient Kathy Yelick of the University of California, Berkeley, discusses the scientific applications fueling exascale computing.
Yelick says these include national security, energy, manufacturing, infrastructure, and scientific discovery, and each field comes with an exascale challenge demanding about 50 times the computational power of current systems.
"These projects are not simply scaling or porting old codes to new machines, but each represents a new predictive or analytic capability," Yelick says.
She sees progress toward exascale, noting, "Much of the [Exascale Computing Program (ECP)] software and programming technology can be leveraged across multiple applications, both within ECP and beyond."
Yelick also expects new machine-learning software will be developed that can support much higher performance levels.
"Some of our policies around the use of [high-performance computing] need to change to better fit data workloads, both to handle on-demand computing for real-time data streams and to address the long-term needs for data provenance and sharing," she says.
View Full Article
Abstracts Copyright © 2017 Information Inc., Bethesda, Maryland, USA