The Compelling Case for Exascale Computing

Information technology and applied science engineering play an essential role in society, from improving decision-making to advancing humanity’s knowledge of the world and the universe. Supercomputing, or high-performance computing (HPC), enables scientists and engineers to push the edge of what is possible for US science and innovation. Using HPC-based modeling and simulation, they are able to study systems that otherwise would be impractical or impossible to investigate in the real world due to their complexity, size, or fleeting nature, or the danger they pose.

Applying the leading-edge capabilities of HPC-based modeling and simulation is essential to the execution of DOE missions in science and engineering and to DOE’s responsibility for stewardship of the nation’s nuclear stockpile. Over the past several decades, sustained technology investment has supported the development of increasingly powerful HPC systems, resulting in leadership status for the United States as well as substantial economic, energy, and national security benefits. But today, as other nations increase their investments in HPC, global competitiveness is on the rise.

To maintain leadership and to address future challenges in economic impact areas and threats to security, the United States is making a strategic move in HPC—a grand convergence of advances in codesign, modeling and simulation, data analytics, machine learning, and artificial intelligence. The success of this convergence hinges on achieving exascale, the next leap forward in computing.

The exponential increase in memory, storage, and compute power made possible by exascale systems will drive breakthroughs in energy production, storage, and transmission; materials science; additive manufacturing; chemical design; artificial intelligence and machine learning; cancer research and treatment; earthquake risk assessment; and many other areas.

With exascale computing, scientists and engineers will be able to solve problems that previously were out of reach, and the effects on the lives of the American people and the world will be profound.

Impact of Exascale Computing

A Milestone in Computing

The fastest supercomputers in the world today solve problems at the petascale, or 1 quadrillion (10¹⁵) operations each second. While these petascale systems are quite powerful, the next milestone in computing achievement, exascale, will be transformative because of the degree of problem-solving capability it will enable—and the benefits in our everyday lives will be far-reaching.

In the most basic sense, exascale is 1,000 times faster and more powerful than petascale. Exascale computing refers to the capability to perform a billion billion (a quintillion) operations per second. The Greek prefix “exa” means 1,000 multiplied by itself six times. Exascale is denoted as 10¹⁸, or as 1 followed by 18 zeros.

At a quintillion operations per second, exascale computers will more realistically simulate the processes involved in scientific discovery and national security such as precision medicine, regional climate, additive manufacturing, the conversion of plants to biofuels, the relationship between energy and water use, the unseen physics in materials discovery and design, and fundamental forces of the universe, and myriad others.

A Strategic Partnership

The DOE-led Exascale Computing Initiative (ECI), a partnership between two DOE organizations, the Office of Science (SC) and the National Nuclear Security Administration (NNSA), was formed in 2016 to accelerate research, development, acquisition, and deployment projects to deliver exascale computing capability to the DOE laboratories by the early to mid-2020s.

Three Major Components of ECI

• Selected program office application development: DOE Office of Science Biological and Environmental Research program and Basic Energy Sciences program and DOE National Nuclear Security Administration
• Exascale system procurement projects and facilities: ALCF-3 (Aurora), OLCF-5 (Frontier), ASC ATS-4 (El Capitan)
• Exascale Computing Project

The ECI consists of three main components: (1) SC and NNSA computer facility site preparation investments, (2) computer vendor nonrecurring engineering activities needed for the delivery of exascale systems within this time frame, and (3) the Exascale Computing Project (ECP), which was launched in 2016 and brings together research, development, and deployment activities as part of a capable exascale computing ecosystem to ensure an enduring exascale computing capability for the nation.

ECP, a 7 year project, is focused on delivering specific applications, software products, and outcomes on DOE computing facilities. Integration across these elements with specific hardware technologies for the manifestation of exascale systems is fundamental to the success of ECP.

The outcome of ECP is the accelerated delivery of a capable exascale computing ecosystem to provide breakthrough solutions that address our most critical challenges in scientific discovery, energy assurance, economic competitiveness, and national security. Thus, the aim is not simply a matter of ensuring more powerful computing systems.

ECP is designed to create more valuable and rapid insights from a wide variety of applications (“capable”), which requires a much higher level of inherent effectiveness in all methods, software tools, and ECP-enabled computing technologies to be acquired by the DOE laboratories (“ecosystem”).

Specifically, advanced leadership computing capabilities are required to

• discover new energy solutions needed for a sustainable future,
• extend our knowledge of the natural world through scientific inquiry,
• maintain a vibrant effort in science and engineering as a cornerstone of the nation’s economic prosperity,
• deliver new technologies to advance DOE’s mission, and
• sustain a world-leading workforce in advanced technology.

ECP’s leadership team has staff from six of the largest DOE national laboratories, but overall, the project has participation from 15 of the 17 DOE laboratories. Today ECP is composed of approximately 1,000 researchers, scientists, participating US HPC systems companies, and project management experts in support of the project’s key research focus areas: Application Development, Software Technology, and Hardware and Integration. ECP will also play a key role in helping to drive new training programs throughout the US HPC ecosystem to prepare application developers, researchers, and scientists to take full advantage of future-generation exascale environments.

ECP Enables US Revolutions in Technology Development: Scientific Discovery; Health Care; and Energy, Economic, and National Security

Looking Ahead

An aggressive research, development, and deployment project, the Exascale Computing Project (ECP) is focused on the delivery of DOE mission-critical applications, an integrated software stack, and exascale hardware technology advances. These products are being deployed to DOE high-performance computing (HPC) facilities on pre-exascale machines and will ultimately be implemented on exascale systems that will address the United States’ most critical challenges in national security, energy assurance, economic competitiveness, health care, and scientific discovery.

Applications

Exascale-capable applications are a foundational element of ECP and will be the delivery vehicle for solutions and insights to key national challenges and emerging technical areas such as machine learning and artificial intelligence. Problems heretofore intractable will be accessible with ECP applications.

Summary of activities:

• Twenty-four Application Development teams have been actively engaged in targeted development and capability enablement for 3 years.
• The applications have been well-defined exascale challenge problem targets with associated “science work rate” goals.
• Initial performance experiences on pre-exascale systems (Summit, Sierra) are exceeding expectations.

Learn more about ECP’s Application Development research focus area.

Software Technologies

Software technologies play an essential enabling role as the underlying technology to application integration and effectiveness on computing systems. An expanded and vertically integrated software stack is being developed to include advanced mathematical libraries and frameworks, extreme-scale programming environments, tools, and I/O and visualization libraries.

Summary of activities:

• The Software Technology research focus area has 35 projects.
• Seventy software technology products are being actively developed for next-generation architectures.
• Regular assessment of software stack products ensures line-of-sight to applications and HPC facilities.
• Delivering products via vendor stacks, community ecosystems such as LLVM, Software Development Kits (SDKs) and the Extreme-scale Scientific Software Stack (E4S), in both source and binary formats, including containers and build caches.

Learn more about ECP’s Software Technology research focus area.

Hardware and Integration

The Hardware and Integration (HI) Area is focused on the deployment and integration of applications, software, and hardware innovations for the forthcoming exascale platforms. Working closely with the DOE HPC facilities, HI helps deliver optimized applications, robust and well tested software, training and allocation and account support on the early hardware and eventual exascale systems.

Summary of activities:

• The PathForward vendor hardware R&D element has completed. The impact of the program is evident by the number of PathForward supported technologies that will be featured on the exascale systems. HI also delivered independent evaluation of hardware technologies.
• Deployment and testing of the E4S software stack on to the early hardware systems and eventual exascale system
• Optimization, porting and tuning of applications to exascale hardware and software environments
• Training for the broad ECP user base
• Allocation and account support for ECP teams running on early hardware and eventual exascale systems

Learn more about ECP’s Hardware and Integration research focus area.