libEnsemble: A Library to Coordinate the Concurrent Evaluation of Dynamic Ensembles of Calculations

Abstract

Almost all applications stop scaling at some point; those that don't are seldom performant when considering time to solution on anything but aspirational/unicorn resources. Recognizing these tradeoffs as well as greater user functionality in a near-term exascale computing era, we present libEnsemble, a library aimed at particular scalability- and capability-stretching uses. libEnsemble enables running concurrent instances of an application in dynamically allocated ensembles through an extensible Python library. We highlight the structure, execution, and capabilities of the library on leading pre-exascale environments as well as advanced capabilities for exascale environments and beyond.

References in corpus (5)

• Parsl: Pervasive Parallel Programming in Python
• Warp-X: a new exascale computing platform for beam-plasma simulations
• Replication or exploration? Sequential design for stochastic simulation experiments
• Porting WarpX to GPU-accelerated platforms
• Breaking (Global) Barriers in Parallel Stochastic Optimization with Wait-Avoiding Group Averaging

Cited by in corpus (7)

• Bayesian optimization of laser-plasma accelerators assisted by reduced physical models
• Energy Compression and Stabilization of Laser-Plasma Accelerators
• Designing a Framework for Solving Multiobjective Simulation Optimization Problems
• Portable, heterogeneous ensemble workflows at scale using libEnsemble
• Optimized matching conditions for self-guided laser wakefield accelerators
• Comparison of Multiobjective Optimization Methods for the LCLS-II Photoinjector
• High brightness, symmetric electron bunch generation in a plasma wakefield accelerator via a radially-polarized plasma photocathode