Exascale Computing Project
Paul Messina is a senior strategic advisor at the ALCF who focuses on future directions for the facility. In 2002-2004, he served as Distinguished Senior Computer Scientist at Argonne and as Adviser to the Director General at CERN (European Organization for Nuclear Research). Previously at Caltech, Dr. Messina served as Director of the Center for Advanced Computing Research, as Assistant Vice President for Scientific Computing, and as Faculty Associate for Scientific Computing. He led the Computational and Computer Science component of Caltech’s research project funded by the Academic Strategic Alliances Program of the Accelerated Strategic Computing Initiative. He also acted as Co-principal Investigator for the National Virtual Observatory and TeraGrid. At Argonne, he held a number of positions from 1973-1987 and was the founding Director of the Mathematics and Computer Science Division.
HPCwire: Hello Paul. Congratulations on being selected as an HPCwire Person to Watch in 2017! Can you provide a brief statement on the mission and execution strategy for the US Exascale Computing Project (ECP)?
Paul Messina: ECP’s mission is to ensure that all the necessary pieces are in place for the first exascale systems – an ecosystem that includes mission critical applications, software stack, hardware architecture, advanced system engineering and hardware components – to enable fully functional, capable exascale computing environments critical to national security, scientific discovery, and a strong U.S. economy.
The ECP develops the strategy, aligns the resources, and conducts the R&D necessary to achieve exascale computing in the United States by 2021. The work is being carried out by staff from DOE national labs, academia, and vendors. Because this is a project led by DOE labs, a leadership team recruited from six of the laboratories manages the activities. I have the honor to be the director of the project, but this is very much a team effort.
The goal is delivered performance to applications, not the number of peak FLOPS. Therefore, we are developing a software stack that will support exascale applications and enable effective use of the exascale hardware. We are funding vendors to do R&D on component technologies and system architectures that will better meet the requirements of applications and will become part of their product roadmaps. Since it will be the supercomputing facilities at the DOE labs procuring the exascale computers (not the ECP), we are collaborating with them to identify the requirements that must be met by the systems. Finally, but perhaps most important, we are supporting the development of full-fledged applications that are important to the missions of the DOE and the National Nuclear Security Administration (NNSA) so that these targeted applications will be ready to use the exascale productively as soon as the systems are available.
Pervasive integration and co-design is the approach we are following to execute the project. We feel it is the only way to ensure that the payoff of these efforts is an endurable and robust exascale ecosystem that will support a broad spectrum of applications and workloads. All the ECP activities work together to identify and implement the features and functionality that will result in exascale systems that will meet the needs of applications and are affordable, manageable, and reliable.
The ECP is a collaborative project of two U.S. Department of Energy organizations, the Office of Science and the National Nuclear Security Administration.
HPCwire: What is the US community doing to ensure a sustainable path to exascale and beyond? Toward that end, what should we be looking out for in 2017?
We are fortunate to have strong interest and support from all segments of the U.S. HPC community. Last summer, we issued an RFP to the U.S. vendor community for hardware R&D and received many strong proposals. In early 2017, we will announce a number of contracts resulting from that RFP and we anticipate issuing a similar RFP later this fiscal year. Sixteen of the DOE national labs are engaged in ECP applications, co-design, software stack development, and evaluation of new architectures. In many cases universities are partners in those activities. One of the application development projects is carried out jointly with the National Cancer Institute of the National Institutes of Health, and we communicate with the Department on Defense on topics of mutual interest.
In FY 2016, we selected 22 applications, 4 co-design centers, and 35 software stack projects, and incorporated some exascale R&D activities that were already underway in DOE and NNSA. In 2017 those projects will begin to tackle the challenges they have targeted. In the process, we will undoubtedly identify gaps in our portfolio and – if funding is available – will solicit and select additional projects to fill the gaps.
Also in 2017 we will kick off the ECP Industry Council, which is an advisory body whose membership is drawn from the HPC industrial user community as well as the commercial HPC software community. This council will guide our activities so that the exascale ecosystem will meet the needs of HPC users in industry.
It is worth noting that the ECP is a collaborative project of two U.S. DOE organizations—the Office of Science and the National Nuclear Security Administration—both of which have been deeply involved in the creation and use of supercomputers since the beginning.
The involvement of all these constituents of the U.S. community is the best possible way to ensure a sustainable path to exascale and beyond.
HPCwire: What critical application areas will benefit from exascale computational capability?
Exascale applications will make major contributions to national security, scientific discovery, and a strong U.S. economy. The topics that have been identified include digital manufacturing, nuclear energy, wind energy, earth system science, weather forecasting, fundamental and theoretical domains like astrophysics and quantum chromodynamics, as well as many aspects of national security.
A few examples are:
- High-efficiency, low-emission combustion engine and gas turbine design
- Reliable and efficient planning of the electric power grid
- Additive manufacturing of qualifiable metal parts
- Oil and gas exploration
- Better prediction of severe weather events
- Cancer research
- Subsurface use for carbon capture, fossil fuel extraction, waste disposal
- Stress-resistant crop analysis and catalytic conversion of biomass-derived alcohols
- Sub-surface modeling in support of carbon cycling, and environmental remediation.
- Design and commercialization of small modular nuclear reactors.
- Biofuel catalyst design
- Seismic hazard risk assessment
While those applications and more will benefit from such exascale capability, in addition the science, methods, and technology that will be spawned along the way will enable new products and new applications that will drive the nation forward in many areas.
To expand on this conversation, here are just a few examples to give a bit more detail:
Combustion science advances that increase efficiency of gas turbines by potentially 25%-50% and lower emissions from internal combustion engines.
Accelerated cancer research through improved understanding of the molecular basis of key protein interactions; developing predictive models for drug response, and automating the analysis; and extraction of information from millions of cancer patient records to determine optimal cancer treatment strategies.
Higher-fidelity simulations for nuclear stockpile stewardship certification and assessments to ensure that the nation’s nuclear weapons stockpile is safe, secure and reliable.
Advanced prediction and qualification of physical structures created through Additive Manufacturing, enabled by advanced simulation of the processes and materials involved.
The design of enzymes and plants optimized for the conversion of biomass to biofuels to relieve U.S. dependence on oil and to facilitate the production of other useful bioproducts
Increased wind plant efficiency through enhanced design of turbine blades as well as placement of turbines in wind farms.
Discovery and characterization of new materials with desired properties such as are needed to manufacture batteries with revolutionary energy and power densities, better stability and safety, faster charging, and longer lifetimes.
Nuclear reactor simulations aimed at safe increased fuel utilization, power upgrades, and reactor life extensions and design of new, safe, cost-effective reactors.
In short, Exascale is a critical technology step in the advancement of U.S. competitiveness with far-reaching socioeconomic benefits.
HPCwire: Outside of the professional sphere, what can you tell us about yourself – personal life, family, background, hobbies, etc.? Is there anything about you your colleagues might be surprised to learn?
Sailing has been my favorite pastime since I was 10 years old. I do not get to sail as much as I would like but it remains my passion. I also enjoy traveling, reading books on a range of topics, and attending live performances of all kinds, including opera, symphony, plays, and music concerts. I love being in nature, walking, hiking, whale watching, and visiting our many beautiful National Parks.
| Guangwen Yang