Topic: Mathematical Optimization

Developed by LLNL, Colorado, and Purdue researchers, a new approach eases the implementation of curved geometries into computing simulations.

The Center for Efficient Exascale Discretizations has developed innovative mathematical algorithms for the DOE’s next generation of supercomputers.

libROM is a library designed to facilitate Proper Orthogonal Decomposition (POD) based Reduced Order Modeling (ROM).

A new component-wise reduced order modeling method enables high-fidelity lattice design optimization.

Researchers will address the challenge of efficiently differentiating large-scale applications for the DOE by building on advances in LLNL’s MFEM finite element library and MIT’s Enzyme AD tool.

Highlights include power grid challenges, performance analysis, complex boundary conditions, and a novel multiscale modeling approach.

As Computing’s fifth Fernbach Fellow, postdoctoral researcher Steven Roberts will develop, analyze, and implement new time integration methods.

Highlights include scalable deep learning, high-order finite elements, data race detection, and reduced order models.

Highlights include debris and shrapnel modeling at NIF, scalable algorithms for complex engineering systems, magnetic fusion simulation, and data placement optimization on GPUs.