skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Embedded ensemble propagation for improving performance, portability, and scalability of uncertainty quantification on emerging computational architectures

Abstract

In this study, quantifying simulation uncertainties is a critical component of rigorous predictive simulation. A key component of this is forward propagation of uncertainties in simulation input data to output quantities of interest. Typical approaches involve repeated sampling of the simulation over the uncertain input data, and can require numerous samples when accurately propagating uncertainties from large numbers of sources. Often simulation processes from sample to sample are similar and much of the data generated from each sample evaluation could be reused. We explore a new method for implementing sampling methods that simultaneously propagates groups of samples together in an embedded fashion, which we call embedded ensemble propagation. We show how this approach takes advantage of properties of modern computer architectures to improve performance by enabling reuse between samples, reducing memory bandwidth requirements, improving memory access patterns, improving opportunities for fine-grained parallelization, and reducing communication costs. We describe a software technique for implementing embedded ensemble propagation based on the use of C++ templates and describe its integration with various scientific computing libraries within Trilinos. We demonstrate improved performance, portability and scalability for the approach applied to the simulation of partial differential equations on a variety of CPU, GPU, and acceleratormore » architectures, including up to 131,072 cores on a Cray XK7 (Titan).« less

Authors:
 [1];  [1];  [1];  [1];  [2];  [1]
  1. Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)
  2. Sandia National Lab. (SNL-CA), Livermore, CA (United States)
Publication Date:
Research Org.:
Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States). Oak Ridge Leadership Computing Facility (OLCF); Sandia National Lab. (SNL-NM), Albuquerque, NM (United States); Sandia National Lab. (SNL-CA), Livermore, CA (United States)
Sponsoring Org.:
USDOE Office of Science (SC); USDOE National Nuclear Security Administration (NNSA)
OSTI Identifier:
1369442
Report Number(s):
SAND-2015-9921J
Journal ID: ISSN 1064-8275; 608080
Grant/Contract Number:  
AC04-94AL85000
Resource Type:
Journal Article: Accepted Manuscript
Journal Name:
SIAM Journal on Scientific Computing
Additional Journal Information:
Journal Volume: 39; Journal Issue: 2; Journal ID: ISSN 1064-8275
Publisher:
SIAM
Country of Publication:
United States
Language:
English
Subject:
97 MATHEMATICS AND COMPUTING; uncertainty quantification; partial differential equations; sparse linear algebra; parallel architectures; vectorization

Citation Formats

Phipps, Eric T., D'Elia, Marta, Edwards, Harold C., Hoemmen, Mark Frederick, Hu, Jonathan J., and Rajamanickam, Sivasankaran. Embedded ensemble propagation for improving performance, portability, and scalability of uncertainty quantification on emerging computational architectures. United States: N. p., 2017. Web. doi:10.1137/15M1044679.
Phipps, Eric T., D'Elia, Marta, Edwards, Harold C., Hoemmen, Mark Frederick, Hu, Jonathan J., & Rajamanickam, Sivasankaran. Embedded ensemble propagation for improving performance, portability, and scalability of uncertainty quantification on emerging computational architectures. United States. https://doi.org/10.1137/15M1044679
Phipps, Eric T., D'Elia, Marta, Edwards, Harold C., Hoemmen, Mark Frederick, Hu, Jonathan J., and Rajamanickam, Sivasankaran. 2017. "Embedded ensemble propagation for improving performance, portability, and scalability of uncertainty quantification on emerging computational architectures". United States. https://doi.org/10.1137/15M1044679. https://www.osti.gov/servlets/purl/1369442.
@article{osti_1369442,
title = {Embedded ensemble propagation for improving performance, portability, and scalability of uncertainty quantification on emerging computational architectures},
author = {Phipps, Eric T. and D'Elia, Marta and Edwards, Harold C. and Hoemmen, Mark Frederick and Hu, Jonathan J. and Rajamanickam, Sivasankaran},
abstractNote = {In this study, quantifying simulation uncertainties is a critical component of rigorous predictive simulation. A key component of this is forward propagation of uncertainties in simulation input data to output quantities of interest. Typical approaches involve repeated sampling of the simulation over the uncertain input data, and can require numerous samples when accurately propagating uncertainties from large numbers of sources. Often simulation processes from sample to sample are similar and much of the data generated from each sample evaluation could be reused. We explore a new method for implementing sampling methods that simultaneously propagates groups of samples together in an embedded fashion, which we call embedded ensemble propagation. We show how this approach takes advantage of properties of modern computer architectures to improve performance by enabling reuse between samples, reducing memory bandwidth requirements, improving memory access patterns, improving opportunities for fine-grained parallelization, and reducing communication costs. We describe a software technique for implementing embedded ensemble propagation based on the use of C++ templates and describe its integration with various scientific computing libraries within Trilinos. We demonstrate improved performance, portability and scalability for the approach applied to the simulation of partial differential equations on a variety of CPU, GPU, and accelerator architectures, including up to 131,072 cores on a Cray XK7 (Titan).},
doi = {10.1137/15M1044679},
url = {https://www.osti.gov/biblio/1369442}, journal = {SIAM Journal on Scientific Computing},
issn = {1064-8275},
number = 2,
volume = 39,
place = {United States},
year = {Tue Apr 18 00:00:00 EDT 2017},
month = {Tue Apr 18 00:00:00 EDT 2017}
}

Journal Article:
Free Publicly Available Full Text
Publisher's Version of Record

Citation Metrics:
Cited by: 6 works
Citation information provided by
Web of Science

Save / Share:

Works referenced in this record:

A Stochastic Collocation Method for Elliptic Partial Differential Equations with Random Input Data
journal, January 2007


On Improving Linear Solver Performance: A Block Variant of GMRES
journal, January 2006


Multi-Jagged: A Scalable Parallel Spatial Partitioning Algorithm
journal, March 2016


Kokkos: Enabling manycore performance portability through polymorphic memory access patterns
journal, December 2014


Polynomial Chaos in Stochastic Finite Elements
journal, March 1990


Stochastic finite element methods for partial differential equations with random input data
journal, May 2014


Latin hypercube sampling and the propagation of uncertainty in analyses of complex systems
journal, July 2003


An overview of the Trilinos project
journal, September 2005


Sparsity: Optimization Framework for Sparse Matrix Kernels
journal, February 2004


Towards Extreme-Scale Simulations for Low Mach Fluids with Second-Generation Trilinos
journal, December 2014


The Monte Carlo Method
journal, September 1949


Quasi-Monte Carlo methods and pseudo-random numbers
journal, January 1978


A Sparse Grid Stochastic Collocation Method for Partial Differential Equations with Random Input Data
journal, January 2008


An Anisotropic Sparse Grid Stochastic Collocation Method for Partial Differential Equations with Random Input Data
journal, January 2008


The block conjugate gradient algorithm and related methods
journal, February 1980


High-Order Collocation Methods for Differential Equations with Random Inputs
journal, January 2005


The Wiener--Askey Polynomial Chaos for Stochastic Differential Equations
journal, January 2002


Works referencing / citing this record:

Prediction and reduction of runtime in non-intrusive forward UQ simulations
journal, August 2019