Static Load Balancing for Multi-level Monte Carlo Finite Volume Solvers
نویسندگان
چکیده
The Multi-Level Monte Carlo finite volumes (MLMC-FVM) algorithm was shown to be a robust and fast solver for uncertainty quantification in the solutions of multi-dimensional systems of stochastic conservation laws. A novel load balancing procedure is used to ensure scalability of the MLMC algorithm on massively parallel hardware. We describe this procedure together with other arising challenges in great detail. Finally, numerical experiments in multi-dimensions showing strong and weak scaling of our implementation are presented.
منابع مشابه
Multi-level Monte Carlo finite volume methods for nonlinear systems of conservation laws in multi-dimensions
We extend the Multi-Level Monte Carlo (MLMC) algorithm of [19] in order to quantify uncertainty in the solutions of multi-dimensional hyperbolic systems of conservation laws with uncertain initial data. The algorithm is presented and several issues arising in the massively parallel numerical implementation are addressed. In particular, we present a novel load balancing procedure that ensures sc...
متن کاملMulti-level Monte Carlo finite volume methods for uncertainty quantification of acoustic wave propagation in random heterogeneous layered medium
We consider the very challenging problem of efficient uncertainty quantification for acoustic wave propagation in a highly heterogeneous, possibly layered, random medium, characterized by possibly anisotropic, piecewise log-exponentially distributed Gaussian random fields. A multi-level Monte Carlo finite volume method is proposed, along with a novel, bias-free upscaling technique that allows t...
متن کاملPartitioning and Load Balancing for Emerging Parallel Applications and Architectures
An important component of parallel scientific computing is partitioning – the assignment of work to processors. This assignment occurs at the start of a computation (“static” partitioning). Often, reassignment also is done during a computation (“dynamic” partitioning) to redistribute work as the computation changes. The goal of partitioning is to assign work to processors in a way that minimize...
متن کاملProbabilistic Multi Objective Optimal Reactive Power Dispatch Considering Load Uncertainties Using Monte Carlo Simulations
Optimal Reactive Power Dispatch (ORPD) is a multi-variable problem with nonlinear constraints and continuous/discrete decision variables. Due to the stochastic behavior of loads, the ORPD requires a probabilistic mathematical model. In this paper, Monte Carlo Simulation (MCS) is used for modeling of load uncertainties in the ORPD problem. The problem is formulated as a nonlinear constrained mul...
متن کاملAutomatic Granularity Control for Load Balancing of Concurrent Particle Simulations
This paper demonstrates the use of automatic gran-ularity control as part of dynamic load balancing for irregular, particle-based simulations. Performance optimization techniques are considered in the context of a concurrent Direct Simulation Monte Carlo method used to study the rareeed gas ow inside three-dimensional plasma reactors. Several computational techniques are used to reduce the over...
متن کامل