Connecting universities, leading-edge science, and meaningful data
By building and maintaining vibrant connections to the academic community, Computing helps LLNL meet its mission of strengthening national security through the development and application of world-class computational science and technology. Computational science evolves so rapidly and along so many fronts that, to define and exploit the leading edge, we engage many academic centers of excellence through joint research, collaborative subcontracts, faculty visits, sabbatical arrangements, student internships, and a postdoctoral research program.
In turn, collaborating with LLNL enables researchers to apply cutting-edge techniques to real-world problems. This process is potentially applicable to a variety of applications, from nuclear weapons effects to efficient manufacturing, from global economics to a basic understanding of the universe. Ideas flow in both directions, and both parties benefit.
Academic collaborations are a well-established avenue for computational and computer science research at LLNL. The Laboratory has been collaborating with academic researchers in this area since its inception, and the number of collaborating institutions and breadth of collaboration areas has grown steadily over the years.
In 2016, Computing participated in more than 300 joint research and collaborative subcontracts with researchers at universities, national laboratories, and government institutions around the world. Projects range from ultrahigh resolution climate simulations to geological data analysis to computational astrophysics to search user interfaces.
While we offer academic collaborators access to leading researchers and computing resources, perhaps our most significant asset is our community of potential users. Developing innovative solutions to many problems in computational science relies heavily on how well the researcher understands the problem domain.
Very often, in fact, our research partners will work directly with scientists who will be using the application or simulation being developed; connecting with users and potential users can generate additional insights into the problem to be solved and applications for the solution.
We put research results to good use at LLNL. Software breakthroughs delivered to applications groups at the Laboratory are often initiated or catalyzed by academic collaborators or visitors. For example, LLNL’s hypre, a scalable solver library, was enhanced with the first provably scalable solver for the positive semidefinite form of Maxwell’s equations on general unstructured meshes as a direct result of a seminar by computational mathematician Jinchao Xu of Penn State.
We build long-term relationships through our collaborations. Some of our joint research projects span many years. Some of our partnerships develop out of student–mentor relationships formed when an academic researcher participated in our postdoctoral or student programs.
And though there are hundreds or even thousands of other institutions pursuing computational research, we are pleased to find that many of our collaborators, once they have worked with us, continue to partner with LLNL for their future research efforts.
- Academic collaborations enable researchers to apply cutting-edge techniques to real-world problems.
- The breadth of academic collaborations with Computing is growing.
- Working directly with potential users at LLNL can shed light on a tough research problem.
CASC Academic Engagement Program
LLNL’s Center for Applied Scientific Computing (CASC) facilitates outreach with the academic research community through the Academic Engagement Program (CASC-AEP). The program includes the Computing Masterworks Lecture Series, support for visiting faculty and collaborators, and assistance with subcontracts.
The Predictive Science Academic Alliance Program (PSAAP) is one example of how we partner with academia. PSAAP is an academic and applied research program that focuses on the emerging field of predictive science, which is the application of verified and validated computational simulations to predict properties and dynamics of complex systems.