Science - Written by on January 26, 2016

ACCEL Program Continues to Experience Growth, Success

Tags: , ,

ACCEL’s industrial projects received their first allocations in 2009 through the OLCF’s Director’s Discretionary (DD) program. In 2015, 24 new industrial projects launched, including projects from firms that were using OLCF for the first time. Image Credit: DOE’s Oak Ridge National Laboratory.

ACCEL’s industrial projects received their first allocations in 2009 through the OLCF’s Director’s Discretionary (DD) program. In 2015, 24 new industrial projects launched, including projects from firms that were using OLCF for the first time. Image Credit: DOE’s Oak Ridge National Laboratory.

Industry program offers opportunities for both large, small companies

As 2016 begins, the Accelerating Competitiveness through Computational Excellence (ACCEL) program at the US Department of Energy’s (DOE’s) Oak Ridge National Laboratory (ORNL) continues to grow and deliver successful results.

ACCEL is an ORNL program that is helping start-up companies to Fortune 500 giants supercharge their competitiveness through access to world-class computational resources and expertise not available elsewhere. These businesses are gaining hands-on experience applying advanced computational tools—accessed through the Oak Ridge Leadership Computing Facility (OLCF), a DOE Office of Science User Facility—to their problems and realizing bottom-line business benefits.

Even as the program marks the completion of its seventh year, Suzy Tichenor, director of the Industrial Partnerships Program, noted that she is asked often to explain why Oak Ridge launched such an effort. She said the answer is threefold.

“First, the program aligns with the ORNL and OLCF missions to produce high impact science and engineering results,” Tichenor said. “Industry conducts high impact science and engineering research but doesn’t have computing systems as powerful as those of the OLCF. By providing access to Titan, we’re enabling these companies to solve more complex problems than they can with their in-house resources.

“Then they must make these results public,” she continued. “That’s generally a requirement. Often companies would not publish this information. But by allowing companies to use the OLCF resources, we are drawing out that information and helping to lift the knowledge base of the broader science and engineering community. So this is a real win–win for everyone.

“And finally, the lab has a mission to foster US competitiveness. As the lab and industry are working together, we are strengthening the nation’s competitiveness infrastructure,” Tichenor said.

In 2015, 24 new industrial projects launched, including projects from firms that were using OLCF for the first time: automotive firm Chrysler-Fiat: rocket manufacturers Orbital ATK and Space-X; industry leading software firms Ansys, Dassault Systemes Simulia, and Numeca; and small software firm Appentra. The other 17 new projects were from companies that already had used OLCF resources successfully and were back with new challenges, including Procter & Gamble, GE, GM, and United Technologies Research Center.

ACCEL’s industrial projects received their first allocations in 2009 through the OLCF’s Director’s Discretionary (DD) program. “Our approach was to think big but take measured steps. There were administrative processes to fine tune, we had to get the word out that ACCEL existed, and companies needed time to put together their initial project ideas,” Tichenor said.

As ACCEL developed a solid foundation, DOE opened the Advanced Scientific Computing Research (ASCR) Leadership Computing Challenge (ALCC) program to industrial researchers.

“It has been encouraging to watch companies grow in their ability to use large-scale systems. A number that began with DD allocations have gone on to compete successfully for larger-scale ALCC allocations,” Tichenor said. “And several of them have progressed to compete successfully for the flagship Innovative and Novel Computational Impact on Theory and Experiment (INCITE) program, which has a high computational readiness bar.

“Thus, our initial hopes were born out. We felt that if companies were given access to our systems along with guidance on how to use them effectively, they would advance in their abilities to apply leadership-level computing to solve more challenging problems,” Tichenor said.

Critical to the program’s success has been the ongoing engagement of the OLCF team. Tjerk Straatsma’s Scientific Computing Group and Ashley Barker and her User Assistance and Outreach team have helped make each industrial project a success. Director of Science Jack Wells frequently engages with the industrial principal investigators to help them think through new approaches to solving their problems.

“It’s not just about cycles on big computers,” Tichenor explained. “The access to OLCF expertise is even more valuable to most of these firms.”

Companies participating in the ACCEL program vary in their industrial focus, ranging from aerospace to consumer products to automotive; they also vary in size, from start-up to multibillion-dollar conglomerates. Tichenor noted small companies can be just as—or sometimes more—experienced than large companies with modeling and simulation.

Success is not based on whether a firm is big or small but rather is directly tied to a firm’s commitment to scaling up.

“We have found very large, global corporations that are inexperienced in modeling and simulation (and not very interested in getting better) and start-up companies, like SmartTruck and KatRisk, that were highly experienced but just needed access to a large-scale system to reach their goals,” Tichenor said.

There are also examples of companies you might not associate with high-performance computing.

“Take, for example, multi-billion dollar insurance company FM Global. It has the largest fire testing experimental facility in the world used to replicate and analyze very large-scale industrial warehouse fires. But even their test facility cannot duplicate the scale of the new mega-warehouses that are springing up,” Tichenor said.

“This team has turned to ramping up its efforts in modeling and simulation to augment and compliment their physical testing. With access to OLCF resources, FM Global’s modeling and simulation skills have grown, and they are simulating larger scale fires than they could on their in-house computers. Modeling and simulation now have become key components for FM Global’s world-leading research into fire growth and suppression.”

These companies represent but a few of the success stories since ACCEL began, and the future appears bright.

“We started modestly and piloted ACCEL using Director’s Discretionary allocations, and it grew from there,” Tichenor said. “Along the way, we and our industrial partners have learned a lot about how to make the process successful. The result has been an ongoing stream of research projects that are producing important new insights for the scientific community and delivering new, successful products to the market.

“We’re excited to see how the next few years unfold.”

 Oak Ridge National Laboratory is supported by the US Department of Energy’s Office of Science. The single largest supporter of basic research in the physical sciences in the United States, the Office of Science is working to address some of the most pressing challenges of our time. For more information, please visit science.energy.gov.