ECP Year 4: Annual Meeting Showcases Project’s Momentum and Milestones
Berkeley Lab Staff Share Exascale Software, Applications, HPC, Networking Expertise
February 7, 2020
More than 40 staff from Berkeley Lab’s Computing Sciences Area (CSA) participated in the Exascale Computing Project’s (ECP) 4th annual meeting, held Feb. 3-6 in Houston, Texas. Established in 2016, ECP is a multi-agency effort within the U.S. Department of Energy (DOE) designed to ensure that the U.S. maintains leadership in high performance computing (HPC) and high-speed networking for science, innovation, and national security.
Some 800 people attended this year’s meeting, including about 600 from 13 DOE national labs and more than 200 from universities, industry, and international collaborators, according to Lori Diachin, ECP Deputy Director. Day 2 kicked off with an enthusiastic keynote by DOE Office of Science Director Chris Fall about the increasing impact of the ECP’s efforts on the future of HPC systems, software, applications, and distributed networks.
“The Exascale Computing Project is about more than just building fast computers,” Fall said. “It’s about coupling the fastest computers to the world’s best instruments for science over the most capable network. It’s about solving bigger problems faster and more accurately.”
Showcasing how much progress has been made over the last four years, the week-long event featured a plethora of tutorials, BoF (Birds of a Feather) sessions, panels, posters, plenary talks, and focus area meetings. “It feels as though the ECP initiative is really hitting its stride,” noted ECP Communications Director Mike Bernhardt, with a number of major accomplishments in the past year and more poised for this year.
Among the 2019 milestones was the much-anticipated Critical Decision-2/3 review, which received high marks across the board, a significant indication that the project is on track to achieve its goals, noted ECP Director Doug Kothe in his “State of the ECP” presentation. (For more details, listen to this new ECP podcast with Kothe.) Other high-level highlights over the past year include:
- Application Development: Finalized metrics for all 24 ECP applications
- Software Technology: Issued an updated Capability Assessment Report in February 2020
- Hardware & Integration: Application performance engineers are now in place at ALCF, OLCF, and NERSC
- Increased outreach activities through the ECP Industry Advisory Council and multiple agencies and facilities in the U.S. and internationally
“Over the past two years, I've been deeply embedded in ECP as Software Technologies Deputy Director and have seen the entire project mature in terms of developing closer connections between applications, software, and the DOE HPC facilities,” said Jonathan Carter, who recently took on the role of Associate Laboratory Director for Berkeley Lab’s Computing Sciences Area. “Everyone will benefit in terms of application performance and portability, software development best practices, and facilities serving their user base more effectively.”
Berkeley Lab’s contributions to these projects continue to be multifaceted as well. Here is a sampling of presentations from the ECP meeting involving Computing Sciences staff:
- Update on AMReX and AMReX Applications – John Bell, Ann Almgren, Kevin Gott, Andrew Myers
- Intelligent Distributed Data Movement for Exascale Computing – co-hosted by ESnet's Chin Guok, with Alex Sim; presentations included overviews of ESnet6 services, SENSE, and results of the xCache pilot study
- OpenMP Roadmap for Accelerators Across DOE Pre-Exascale/Exascale Machines – Chris Daley, Jack Deslippe, Brian Friesen, Jay Srinivisan
- EQSIM: Transforming Earthquake Hazard and Risk Assessment Through Exascale Simulations – David McCallen
- UPC++: A PGAS/RPC Library for Asynchronous Exascale Communication in C++ – Amir Kamil
- Tools/Techniques for Application Performance Characteristics – Sam Williams
- ECP Continuous Integration Startup Tutorial – Tiffany Connors
- Container Utilization at DOE Compute Facilities – Shane Canon
- Application Integration in the User Facilities: Lessons Learned so Far – Debbie Bard
- Perlmutter – A Waypoint for ECP Teams: Brandon Cook, Brian Friesen, Jack Deslippe, Jay Srinivisan
- Exascale I/O with ADIOS – John Wu, Jin Min Gu
- Early Experience of Application Developers with OpenMP Offloading – Chris Daley, Doug Doerfler, Brian Friesen, Hadia Ahmed
- ExaGraph – Aydin Bulic
- Advances in Science and Engineering Enabled by ECP Applications – Jean-Luc Vay
- Achieving High-Performance I/O with HDF5 – Quincey Koziol
- STRUMPACK/SuperLu: Fast Parallel Direct Linear Solvers and Preconditioners – Sherry Li, Lang Liu
- System Software: Evaluation and Future for Pre-Exascale and Exascale Systems – Doug Jacobsen
In addition to ECP’s support of these and other R&D efforts, the organization works closely with a number of vendors who are actively adopting HPC into their design and manufacturing processes and developing new HPC technologies. These advances were showcased at the meeting’s Industry Panel, organized by ECP’s Industry Advisory Council, where four vendor partners discussed “Industry Imperatives for Exascale”: moderator Dave Kepcznyski, CIO of GE Research; Prith Banerjee, CTO of ANSYS; Frank Ham, CEO of Cascade Technologies; and Bill Nitzberg, CTO of PBS Works at Altair Engineering. They shared how their companies and customers are using HPC, what challenges remain in getting some industries to fully embrace new design and manufacturing technologies, and their excitement as exascale computing capabilities move steadily toward reality.
“The focus is no longer just on applications and software development but around the researchers,” Kepcznyski said, echoing the sentiments Dr. Fall expressed in his keynote address. “It is about building an ecosystem that is reliable, sustainable, and portable.”
About Computing Sciences at Berkeley Lab
High performance computing plays a critical role in scientific discovery. Researchers increasingly rely on advances in computer science, mathematics, computational science, data science, and large-scale computing and networking to increase our understanding of ourselves, our planet, and our universe. Berkeley Lab’s Computing Sciences Area researches, develops, and deploys new foundations, tools, and technologies to meet these needs and to advance research across a broad range of scientific disciplines.
Founded in 1931 on the belief that the biggest scientific challenges are best addressed by teams, Lawrence Berkeley National Laboratory and its scientists have been recognized with 13 Nobel Prizes. Today, Berkeley Lab researchers develop sustainable energy and environmental solutions, create useful new materials, advance the frontiers of computing, and probe the mysteries of life, matter, and the universe. Scientists from around the world rely on the Lab’s facilities for their own discovery science. Berkeley Lab is a multiprogram national laboratory, managed by the University of California for the U.S. Department of Energy’s Office of Science.
DOE’s Office of Science is the single largest supporter of basic research in the physical sciences in the United States, and is working to address some of the most pressing challenges of our time. For more information, please visit energy.gov/science.