DOE Basic Energy Sciences Program Contemplating Its Role in Advanced Computing

Share This

Share/Save
Publication date: 
18 July 2017
Number: 
97

At last week’s meeting of the Basic Energy Sciences Advisory Committee, participants discussed the Department of Energy Office of Science’s work in advanced computing. They also discussed research priorities identified at two recent workshops and a new charge to commemorate the past 40 years of work in DOE’s Basic Energy Sciences program.

At its meeting last week, the Basic Energy Sciences Advisory Committee (BESAC) discussed one of the Department of Energy’s top priorities: advanced computing. While DOE’s Basic Energy Sciences (BES) program is not at the core of the department’s advanced computing efforts, those efforts have such broad reach that BES is carefully considering their long-term implications for its work.

Advanced computing a bipartisan and DOE-wide priority

Momentum behind advanced computing has been building within the federal government for some years. The Obama administration supported efforts such as DOE’s Exascale Initiative, and two years ago it launched the National Strategic Computing Initiative, a concerted interagency effort to expedite progress in vanguard computing technologies. The Trump administration has also made advanced computing a major priority. Notably, although it proposes to cut funding for the DOE Office of Science by 17 percent, it has proposed a 12 percent increase for the office’s Advanced Scientific Computing Research (ASCR) program.

BES’s enacted budget for fiscal year 2017 provides over $13 million for a new Computational Chemical Sciences Program, which aims to fund projects that will take advantage of petascale and future exascale capabilities. The program has just completed its first solicitation of applications for funding. At the BESAC meeting, BES Director Harriet Kung reported that DOE is planning to continue the program in future years at the same funding level.

BES is also considering other ways that BES-supported research could exploit advanced computing capabilities. In November 2015, it held a workshop, “BES Computing and Data Requirements in the Exascale Age,” and released a summary report on it earlier this year.

Meanwhile, the Office of Science has initiated a sweeping review of “opportunities and challenges for future high performance computing capabilities,” with a report due near the end of this year.

The increasing seriousness of planning for beyond the exascale program has raised the question of how basic research programs can not only benefit from but also contribute to progress in computing once semiconductor transistors reach their inherent limits. Addressing this question, Kung noted that BES nanoscale and quantum materials and chemistry programs are integral to R&D programs in the development of various quantum devices.

besac-kung-quantum-science.jpg


A slide from BES Director Harriet Kung’s presentation to BESAC showing how BES research and capabilities contribute to quantum information science conducted in other Office of Science programs.

(Image credit – DOE)

However, the full scope and shape of future efforts is far from fully determined. To obtain expert perspectives, BESAC heard presentations from Barbara Helland, associate director of ASCR; Carl Williams, the acting director of the National Institute of Standards and Technology’s Physical Measurement Laboratory; and Stan Williams of Hewlett-Packard. Together, they offered an enthusiastic picture of the future of quantum and neuromorphic computing and other technologies that could not only increase performance but also yield revolutionary gains in energy efficiency.

Asked by BESAC Chair Persis Drell how BES might structure its efforts in the area, Helland urged the program to remain informed about developments in computing because of the rapid rate of progress being made. Stan Williams suggested that computation science should itself be supported as a field of fundamental scientific inquiry. “It’s a true area of science, as demonstrated by [Alan] Turing and the greats of 70 years ago, and it’s one that we’ve just sort of dropped,” he said.

BES prioritizing for both short and long term

While BES becomes more deeply involved with advanced computing, it is also considering its priorities in other areas. Most immediately, like many other federal science programs, it has had to make plans in view of the Trump administration’s proposed budget cuts.

Kung reported that the budget calls for the closure of the Batteries and Energy Storage and the Fuels from Sunlight Energy Innovation Hubs as well as Center for Functional Nanomaterials and the Center for Integrated Nanotechnologies. In addition, selected light source beamlines and neutron flight paths would be shut down. The Stanford Synchrotron Radiation Light Source would only operate for one quarter of the year. Research funding would decrease by 18 percent.

However, Kung noted that the House has proposed flat funding for BES in its appropriations bill. The bill includes a 4 percent cut in BES’s research budget and no funding for energy innovation hubs, though it does provide $10 million to continue similar research. It also provides construction funding for the Advanced Photon Source upgrade at Argonne National Laboratory that is $47 million higher than the administration’s $20 million request. Senate appropriators had not released appropriations legislation at the time of the meeting, but it is now known that they are seeking a 3 percent increase for the Office of Science as a whole.

Looking to the longer term, BESAC also heard updates from two Basic Research Needs (BRN) workshops held earlier this year on next generation electrical energy storage and catalysis science, respectively. Full reports from the workshops are forthcoming, but the priority research directions identified at the workshops are detailed in the presentations’ slide decks, available here and here. Another BRN workshop will be held August 8 and 9 on research needs related to future nuclear energy.

BESAC to promote 40 years of scientific achievement

Drell also announced BESAC’s first new charge since December 2015: to prepare a report commemorating the 40th anniversary of the BES program.

In the charge letter, Steve Binkley, the acting director of the Office of Science, wrote, “The report should highlight a few outstanding examples of major scientific accomplishments emerging from BES support that have shaped the fields of BES research, with an eye toward learning from these examples to motivate BES investment strategies for the future.”

Drell said that physicist and BESAC member Marc Kastner will chair the subcommittee responsible for the charge. She also said that she envisions the final product as a brief and attractively illustrated document that would tell compelling stories justifying investment in basic science. It would be directed toward a broad audience, “from congressional staffers to the educated public.” She emphasized, “It is not for us around the table, nor for any of us in this room.”

Discussing the charge, BESAC members began to address the problem of how to select particular accomplishments that would best convey BES’s success. Questions included whether the report should concentrate only on scientific discoveries or also on the tools that enabled them, and whether one should choose projects that show the importance of consistent, long-term support.

Drell said that she regards it as “a very important charge and a very important opportunity for BESAC to contribute to BES.”

About the author

headshot of Will Thomas
wthomas [at] aip.org
+1 301-209-3097