An extreme-scale implicit solver for complex PDEs: Highly heterogeneous flow in earth's mantle

Johann Rudi, A. Cristiano I Malossi, Tobin Isaac, Georg Stadler, Michael Gurnis, Peter W J Staar, Yves Ineichen, Costas Bekas, Alessandro Curioni, Omar Ghattas

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Mantle convection is the fundamental physical process within earth's interior responsible for the thermal and geological evolution of the planet, including plate tectonics. The mantle is modeled as a viscous, incompressible, non-Newtonian fluid. The wide range of spatial scales, extreme variability and anisotropy in material properties, and severely nonlinear rheology have made global mantle convection modeling with realistic parameters prohibitive. Here we present a new implicit solver that exhibits optimal algorithmic performance and is capable of extreme scaling for hard PDE problems, such as mantle convection. To maximize accuracy and minimize runtime, the solver incorporates a number of advances, including aggressive multi-octree adaptivity, mixed continuous-discontinuous discretization, arbitrarily-high-order accuracy, hybrid spectral/geometric/algebraic multigrid, and novel Schur-complement preconditioning. These features present enormous challenges for extreme scalability. We demonstrate that -contrary to conventional wisdom - -algorithmically optimal implicit solvers can be designed that scale out to 1.5 million cores for severely nonlinear, ill-conditioned, heterogeneous, and anisotropic PDEs.

Original languageEnglish (US)
Title of host publicationProceedings of SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis
PublisherIEEE Computer Society
Volume15-20-November-2015
ISBN (Print)9781450337236
DOIs
StatePublished - Nov 15 2015
EventInternational Conference for High Performance Computing, Networking, Storage and Analysis, SC 2015 - Austin, United States
Duration: Nov 15 2015Nov 20 2015

Other

OtherInternational Conference for High Performance Computing, Networking, Storage and Analysis, SC 2015
CountryUnited States
CityAustin
Period11/15/1511/20/15

Fingerprint

Earth (planet)
Planets
Tectonics
Rheology
Scalability
Materials properties
Anisotropy
Fluids
Convection
Hot Temperature

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Computer Science Applications
  • Hardware and Architecture
  • Software

Cite this

Rudi, J., Malossi, A. C. I., Isaac, T., Stadler, G., Gurnis, M., Staar, P. W. J., ... Ghattas, O. (2015). An extreme-scale implicit solver for complex PDEs: Highly heterogeneous flow in earth's mantle. In Proceedings of SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis (Vol. 15-20-November-2015). [a5] IEEE Computer Society. https://doi.org/10.1145/2807591.2807675

An extreme-scale implicit solver for complex PDEs : Highly heterogeneous flow in earth's mantle. / Rudi, Johann; Malossi, A. Cristiano I; Isaac, Tobin; Stadler, Georg; Gurnis, Michael; Staar, Peter W J; Ineichen, Yves; Bekas, Costas; Curioni, Alessandro; Ghattas, Omar.

Proceedings of SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis. Vol. 15-20-November-2015 IEEE Computer Society, 2015. a5.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Rudi, J, Malossi, ACI, Isaac, T, Stadler, G, Gurnis, M, Staar, PWJ, Ineichen, Y, Bekas, C, Curioni, A & Ghattas, O 2015, An extreme-scale implicit solver for complex PDEs: Highly heterogeneous flow in earth's mantle. in Proceedings of SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis. vol. 15-20-November-2015, a5, IEEE Computer Society, International Conference for High Performance Computing, Networking, Storage and Analysis, SC 2015, Austin, United States, 11/15/15. https://doi.org/10.1145/2807591.2807675
Rudi J, Malossi ACI, Isaac T, Stadler G, Gurnis M, Staar PWJ et al. An extreme-scale implicit solver for complex PDEs: Highly heterogeneous flow in earth's mantle. In Proceedings of SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis. Vol. 15-20-November-2015. IEEE Computer Society. 2015. a5 https://doi.org/10.1145/2807591.2807675
Rudi, Johann ; Malossi, A. Cristiano I ; Isaac, Tobin ; Stadler, Georg ; Gurnis, Michael ; Staar, Peter W J ; Ineichen, Yves ; Bekas, Costas ; Curioni, Alessandro ; Ghattas, Omar. / An extreme-scale implicit solver for complex PDEs : Highly heterogeneous flow in earth's mantle. Proceedings of SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis. Vol. 15-20-November-2015 IEEE Computer Society, 2015.
@inproceedings{a9f4b53729b241d0887c460d7c7461bc,
title = "An extreme-scale implicit solver for complex PDEs: Highly heterogeneous flow in earth's mantle",
abstract = "Mantle convection is the fundamental physical process within earth's interior responsible for the thermal and geological evolution of the planet, including plate tectonics. The mantle is modeled as a viscous, incompressible, non-Newtonian fluid. The wide range of spatial scales, extreme variability and anisotropy in material properties, and severely nonlinear rheology have made global mantle convection modeling with realistic parameters prohibitive. Here we present a new implicit solver that exhibits optimal algorithmic performance and is capable of extreme scaling for hard PDE problems, such as mantle convection. To maximize accuracy and minimize runtime, the solver incorporates a number of advances, including aggressive multi-octree adaptivity, mixed continuous-discontinuous discretization, arbitrarily-high-order accuracy, hybrid spectral/geometric/algebraic multigrid, and novel Schur-complement preconditioning. These features present enormous challenges for extreme scalability. We demonstrate that -contrary to conventional wisdom - -algorithmically optimal implicit solvers can be designed that scale out to 1.5 million cores for severely nonlinear, ill-conditioned, heterogeneous, and anisotropic PDEs.",
author = "Johann Rudi and Malossi, {A. Cristiano I} and Tobin Isaac and Georg Stadler and Michael Gurnis and Staar, {Peter W J} and Yves Ineichen and Costas Bekas and Alessandro Curioni and Omar Ghattas",
year = "2015",
month = "11",
day = "15",
doi = "10.1145/2807591.2807675",
language = "English (US)",
isbn = "9781450337236",
volume = "15-20-November-2015",
booktitle = "Proceedings of SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis",
publisher = "IEEE Computer Society",

}

TY - GEN

T1 - An extreme-scale implicit solver for complex PDEs

T2 - Highly heterogeneous flow in earth's mantle

AU - Rudi, Johann

AU - Malossi, A. Cristiano I

AU - Isaac, Tobin

AU - Stadler, Georg

AU - Gurnis, Michael

AU - Staar, Peter W J

AU - Ineichen, Yves

AU - Bekas, Costas

AU - Curioni, Alessandro

AU - Ghattas, Omar

PY - 2015/11/15

Y1 - 2015/11/15

N2 - Mantle convection is the fundamental physical process within earth's interior responsible for the thermal and geological evolution of the planet, including plate tectonics. The mantle is modeled as a viscous, incompressible, non-Newtonian fluid. The wide range of spatial scales, extreme variability and anisotropy in material properties, and severely nonlinear rheology have made global mantle convection modeling with realistic parameters prohibitive. Here we present a new implicit solver that exhibits optimal algorithmic performance and is capable of extreme scaling for hard PDE problems, such as mantle convection. To maximize accuracy and minimize runtime, the solver incorporates a number of advances, including aggressive multi-octree adaptivity, mixed continuous-discontinuous discretization, arbitrarily-high-order accuracy, hybrid spectral/geometric/algebraic multigrid, and novel Schur-complement preconditioning. These features present enormous challenges for extreme scalability. We demonstrate that -contrary to conventional wisdom - -algorithmically optimal implicit solvers can be designed that scale out to 1.5 million cores for severely nonlinear, ill-conditioned, heterogeneous, and anisotropic PDEs.

AB - Mantle convection is the fundamental physical process within earth's interior responsible for the thermal and geological evolution of the planet, including plate tectonics. The mantle is modeled as a viscous, incompressible, non-Newtonian fluid. The wide range of spatial scales, extreme variability and anisotropy in material properties, and severely nonlinear rheology have made global mantle convection modeling with realistic parameters prohibitive. Here we present a new implicit solver that exhibits optimal algorithmic performance and is capable of extreme scaling for hard PDE problems, such as mantle convection. To maximize accuracy and minimize runtime, the solver incorporates a number of advances, including aggressive multi-octree adaptivity, mixed continuous-discontinuous discretization, arbitrarily-high-order accuracy, hybrid spectral/geometric/algebraic multigrid, and novel Schur-complement preconditioning. These features present enormous challenges for extreme scalability. We demonstrate that -contrary to conventional wisdom - -algorithmically optimal implicit solvers can be designed that scale out to 1.5 million cores for severely nonlinear, ill-conditioned, heterogeneous, and anisotropic PDEs.

UR - http://www.scopus.com/inward/record.url?scp=84966606327&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84966606327&partnerID=8YFLogxK

U2 - 10.1145/2807591.2807675

DO - 10.1145/2807591.2807675

M3 - Conference contribution

AN - SCOPUS:84966606327

SN - 9781450337236

VL - 15-20-November-2015

BT - Proceedings of SC 2015: The International Conference for High Performance Computing, Networking, Storage and Analysis

PB - IEEE Computer Society

ER -