The U.S. Department of Energy’s (DOE) Innovative and Novel Computational Impact on Theory and Experiment (INCITE) program is the primary means by which researchers gain access to the leadership-class supercomputers at DOE’s Argonne and Oak Ridge National Laboratories.
INCITE’s annual call for proposals, which is now open through June 18, 2021, provides an opportunity for researchers to pursue transformational advances in science and technology through large allocations of computer time and supporting resources at the Argonne Leadership Computing Facility (ALCF) and the Oak Ridge Leadership Computing Facility (OLCF), both DOE Office of Science User Facilities.
This year, Katherine Riley, director of science at the ALCF, took the reins as the new INCITE program manager. In the Q&A below, Riley discusses the aim of the program, its new early career track and how researchers can prepare their science for DOE’s powerful supercomputers.
Q: What is the goal of the INCITE program?
INCITE was created to support high-impact, novel science campaigns that need the biggest computing and data resources in the world. It enables creative and innovative research across different fields of science and engineering that could not be done anywhere else. That’s really why we’re here — to help move scientific computing forward in exciting new directions.
Q: Who can apply for the program?
Anyone can apply for INCITE. That means you can be at a university, you can be in another country, you can work for a company. It is open to everyone. But beyond that, what we’re looking for are people who are having difficulty finding the appropriate computing resources to tackle their computational or data challenge. The idea behind INCITE is you’re getting awards that are substantively bigger than you would get at other computing facilities, and you’re also getting access to the in-house experts at ALCF and OLCF to make sure you’re able to get the most from our supercomputers. So, we’re looking for those researchers who can’t advance their science without that larger capability that you get at a leadership computing facility.
Q: What is INCITE’s new Early Career Track?
For this year’s call, we have committed 10% of INCITE time at ALCF and OLCF to early career researchers. To be eligible, you need to be within 10 years of getting your Ph.D. The proposals will still go through the same review process, but the meritorious submissions from PIs (principal investigators) early in their career will be considered separately as part of this track.
The motivation for this is to support and help cultivate the next generation of computational scientists. We have seen early career researchers be successful in INCITE, but it can be daunting to jump into this highly competitive program full of renowned scientists who have been through this process before. We want to give early career researchers a landing spot, so it doesn’t feel like they have to jump into the middle of the ocean at the very start. INCITE is very competitive but it is worthwhile getting your feet wet in the proposal process. Even if you’re not successful in your first year, you can learn from the process, take the feedback and come back with an even stronger proposal next year.
Q: What supercomputers are available through INCITE?
For the past several years, we’ve had two primary resources in the pool. There is Summit, a 200-petaflop, GPU-accelerated system at Oak Ridge, and Theta, a 12-petaflop, CPU-based system at Argonne. But starting in 2022, for the people planning to apply for INCITE this year, Argonne will also have a new GPU-accelerated resource called Polaris.
And looking ahead to 2023, INCITE projects will be among the first to have access to DOE’s upcoming exascale systems, Aurora at Argonne and Frontier at Oak Ridge. So, it’s an exciting time to apply for an INCITE project. We’ve been in a period of stable resources for a long time, but we’re going to have a big explosion of new capabilities over the next two years with three powerful new systems hitting the floor.
Q: INCITE proposals are reviewed for both scientific merit and computational readiness. Any advice on how researchers can get computationally ready for DOE’s leadership-class systems?
The best thing you can do is request a discretionary allocation at the facility that you’re interested in using. Both ALCF and OLCF have online forms you can fill out to request a modest discretionary award to test your applications, your workflow and your infrastructure. We strongly recommend this for projects that are looking to apply for INCITE. Testing on the targeted systems improves your computational readiness which increases your competitiveness. We just ask researchers to be transparent about what work is ready to go and what work needs further testing or development. The more we understand about what you need to do to get your project moving forward, the more confidence we’ll have in the readiness of your proposal and the more knowledge we’ll have about how we can help you succeed.
There can also be value in using other large systems that may have similar characteristics to the LCF systems. While they won’t be exactly the same, they can provide some confidence that your application is scalable and that your approach is right. But I do encourage people to try to get on the actual LCF resources they want to use because every one of these big systems is different.
Q: The INCITE program has evolved in recent years. Can you tell us about the move to support research involving data-intensive computing, artificial intelligence and machine learning in addition to traditional modeling and simulation campaigns?
The shift to support more data-intensive and machine learning projects is being driven by the scientists and researchers we work with. For many researchers, the way they conduct science is evolving. We’ve been seeing a dramatic increase from one year to another in proposals that have some components of simulation, data and learning. And our facilities have been building up the capabilities and expertise to support this type of research for years now.
As a program, we’re trying to make sure that this change is represented in how we review the proposals. The questions we might ask if you’re doing traditional modeling and simulation might be different than if you’re doing a machine learning-driven project. From the program side, we’ve adjusted our review panels to make sure that we have experts who can speak to how machine learning applies to a specific type of problem and we use that in our assessment. This is a natural step to ensure we are fully informed about what a proposal’s impact might be.
It’s going to be another exciting year for INCITE. We’re looking forward to seeing a new set of ideas for using DOE supercomputers to propel scientific discoveries and innovation.
For an opportunity to ask your own questions about the INCITE program, register for the INCITE proposal writing webinar on June 4, 2021.
The Argonne Leadership Computing Facility provides supercomputing capabilities to the scientific and engineering community to advance fundamental discovery and understanding in a broad range of disciplines. Supported by the U.S. Department of Energy’s (DOE’s) Office of Science, Advanced Scientific Computing Research (ASCR) program, the ALCF is one of two DOE Leadership Computing Facilities in the nation dedicated to open science.
Argonne National Laboratory seeks solutions to pressing national problems in science and technology. The nation’s first national laboratory, Argonne conducts leading-edge basic and applied scientific research in virtually every scientific discipline. Argonne researchers work closely with researchers from hundreds of companies, universities, and federal, state and municipal agencies to help them solve their specific problems, advance America’s scientific leadership and prepare the nation for a better future. With employees from more than 60 nations, Argonne is managed by UChicago Argonne, LLC for the U.S. Department of Energy’s Office of Science.
The U.S. Department of Energy’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, visit https://energy.gov/science.