Rob Neely Makes Connections
Livermore computer scientist and high performance computing (HPC) advocate Rob Neely relishes the diverse nature of his responsibilities. “I like the breadth of work here,” he says. “I also like to help build connections and see them blossom into collaborations. That’s very satisfying.”
Rob currently leads several initiatives for the Advanced Simulation and Computing (ASC) program. For instance, he assists code teams in developing strategies to ensure that their massive physics codes perform optimally on advanced and potentially novel computer architectures. He also coordinates the program’s co-design efforts. LLNL has a long and successful history of design collaboration with hardware vendors, but co-design offers a model for developing deeper and more enduring partnerships between hardware and software experts. Says Rob, “Co-design is about where the state of computing is going, rather than just focusing on creating one specific machine.”
Rob holds additional leadership positions within Livermore Computing and the Center for Applied Scientific Computing (CASC). For his Livermore Computing role, Rob is part of the multi-laboratory team crafting the technical requirements for LLNL’s next-generation supercomputer, to be delivered in 2017. He also works with thought leaders at LLNL, the Department of Energy, and industry to build local and national supercomputing strategies for reaching exascale and beyond. As CASC’s associate division leader, he is helping staff members who wish to work jointly as university researchers or professors to navigate policies and processes and to build win–win relationships with academic partners.
Rob’s diverse yet interrelated assignments demand excellence at fostering communication and connections. Fortunately, he had a chance to hone these skills not so long ago while “on loan” to the Council on Competitiveness, where he worked to reinvigorate the nonpartisan policy organization’s efforts to encourage companies to adopt HPC. During his year in Washington, Rob frequently met with stakeholders in the manufacturing industry to learn about their business needs and explain HPC’s benefits. “If big companies can incorporate modeling and simulation into their supply chain, they can innovate more intelligently and a lot faster,” Rob observes. Meanwhile, Rob learned to communicate more effectively with nonscientists about technical topics and realized that “it really takes a lot of time and effort to make collaborations happen and grow.”
Rob’s interest in HPC began during his undergraduate years, at a supercomputing summer program he attended at Cornell University. After graduate school, Rob joined the newly established Accelerated Strategic Computing Initiative (ASCI, since renamed ASC) as one of the program’s first Livermore hires. The multi-laboratory program analyzes and certifies the safety and reliability of nuclear weapons using modeling and simulation, rather than testing. Rob spent his first fifteen years at LLNL on a multidisciplinary ASCI team tasked with maintaining and enhancing the multiphysics code ALE3D. This experience, perhaps more than any other, convinced him of HPC’s value. “Through my involvement in ASCI, I saw how HPC could accelerate science. Supercomputing has truly changed the way science is done,” says Rob.
Imagine computers so powerful and intricately programmed that they allow us to understand the impacts of explosives without blowing anything up, optimize a product’s design before the first metal piece is forged, or unravel the physics mysteries at a star’s center. Making that vision a reality requires the efforts of thousands. I am part of a vast nationwide team working to build the hardware and software ecosystem that allows us to push the frontiers of computing while also making it approachable to a broader audience.
I also recently started playing the piano for fun.