Lessons Learned from Optimizing Kernels for Adaptive Aggregation Multi-grid Solvers in Lattice QCD

Bálint Joó; Thorsten Kurth

doi:10.1007/978-3-030-02465-9_34

Lessons Learned from Optimizing Kernels for Adaptive Aggregation Multi-grid Solvers in Lattice QCD

Bálint Joó, Thorsten Kurth

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

Abstract

In recent years, adaptive aggregation multi-grid (AAMG) methods have become the gold standard for solving the Dirac equation in Lattice QCD (LQCD) using Wilson-Clover fermions. These methods are able to overcome the critical slowing down as quark masses approach their physical values and are thus the go-to method for performing Lattice QCD calculations at realistic physical parameters. In this paper we discuss the optimization of a specific building block for implementing AAMG for Wilson-Clover fermions from LQCD, known as the coarse restrictor operator, on contemporary Intel processors featuring large SIMD widths and high thread counts. We will discuss in detail the efficient use of OpenMP and Intel vector intrinsics in our attempts to exploit fine grained parallelism on the coarsest levels. We present performance optimizations and discuss the ramifications for implementing a full AAMG stack on Intel Xeon Phi Knights Landing and Skylake processors.

Original language	English
Title of host publication	High Performance Computing - ISC High Performance 2018 International Workshops, Revised Selected Papers
Editors	Rio Yokota, Michèle Weiland, Sadaf Alam, John Shalf
Publisher	Springer Verlag
Pages	472-486
Number of pages	15
ISBN (Print)	9783030024642
DOIs	https://doi.org/10.1007/978-3-030-02465-9_34
State	Published - 2018
Externally published	Yes
Event	International Conference on High Performance Computing, ISC High Performance 2018 - Frankfurt, Germany Duration: Jun 28 2018 → Jun 28 2018

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11203 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	International Conference on High Performance Computing, ISC High Performance 2018
Country/Territory	Germany
City	Frankfurt
Period	06/28/18 → 06/28/18

Funding

Acknowledgment. This research used resources of the National Energy Research Scientific Computing Center (NERSC), a DOE Office of Science User Facility supported by the Office of Science of the U.S. Department of Energy under Contract No. DE-AC02-05CH11231, and of the ALCF, which is supported by DOE/SC under contract DE-AC02-06CH11357. B. Joo acknowledges funding from the DOE Office Of Science, Offices of Nuclear Physics and Advanced Scientific Computing Research through the SciDAC program. B. Joo also acknowledges support from the U.S. DOE Exascale Computing Project (ECP). This work is supported by the U.S. Department of Energy, Office of Science, Office of Nuclear Physics under contract DE-AC05-06OR23177. B. Joo would like to thank and acknowledge Kate Clark of NVIDIA for many discussions about expressing and mapping parallelism in multi-grid solver components in a variety of programming models and hardware and her helpful comments after a reading of this manuscript, as well as Christian Trott of Sandia Labs for discussions about nested par-alleism in OpenMP. This work used resources provided by the Performance Research Laboratory at the University of Oregon. We would especially like to thank Sameer Shende and Rob Yelle for their professional support of the Performance Research Laboratory computers and their timely response to our requests.

Funders	Funder number
DOE Office of Science
DOE/SC	DE-AC02-06CH11357
Office of Nuclear Physics	DE-AC05-06OR23177
U.S. DOE
U.S. Department of Energy	DE-AC02-05CH11231
Office of Science
Advanced Scientific Computing Research

Access to Document

10.1007/978-3-030-02465-9_34

Cite this

Joó, B., & Kurth, T. (2018). Lessons Learned from Optimizing Kernels for Adaptive Aggregation Multi-grid Solvers in Lattice QCD. In R. Yokota, M. Weiland, S. Alam, & J. Shalf (Eds.), High Performance Computing - ISC High Performance 2018 International Workshops, Revised Selected Papers (pp. 472-486). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11203 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-030-02465-9_34

Joó, Bálint ; Kurth, Thorsten. / Lessons Learned from Optimizing Kernels for Adaptive Aggregation Multi-grid Solvers in Lattice QCD. High Performance Computing - ISC High Performance 2018 International Workshops, Revised Selected Papers. editor / Rio Yokota ; Michèle Weiland ; Sadaf Alam ; John Shalf. Springer Verlag, 2018. pp. 472-486 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{68de5440c287430c96cf42bda2501684,

title = "Lessons Learned from Optimizing Kernels for Adaptive Aggregation Multi-grid Solvers in Lattice QCD",

abstract = "In recent years, adaptive aggregation multi-grid (AAMG) methods have become the gold standard for solving the Dirac equation in Lattice QCD (LQCD) using Wilson-Clover fermions. These methods are able to overcome the critical slowing down as quark masses approach their physical values and are thus the go-to method for performing Lattice QCD calculations at realistic physical parameters. In this paper we discuss the optimization of a specific building block for implementing AAMG for Wilson-Clover fermions from LQCD, known as the coarse restrictor operator, on contemporary Intel processors featuring large SIMD widths and high thread counts. We will discuss in detail the efficient use of OpenMP and Intel vector intrinsics in our attempts to exploit fine grained parallelism on the coarsest levels. We present performance optimizations and discuss the ramifications for implementing a full AAMG stack on Intel Xeon Phi Knights Landing and Skylake processors.",

author = "B{\'a}lint Jo{\'o} and Thorsten Kurth",

note = "Publisher Copyright: {\textcopyright} 2018, Springer Nature Switzerland AG.; International Conference on High Performance Computing, ISC High Performance 2018 ; Conference date: 28-06-2018 Through 28-06-2018",

year = "2018",

doi = "10.1007/978-3-030-02465-9_34",

language = "English",

isbn = "9783030024642",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

pages = "472--486",

editor = "Rio Yokota and Mich{\`e}le Weiland and Sadaf Alam and John Shalf",

booktitle = "High Performance Computing - ISC High Performance 2018 International Workshops, Revised Selected Papers",

}

Joó, B & Kurth, T 2018, Lessons Learned from Optimizing Kernels for Adaptive Aggregation Multi-grid Solvers in Lattice QCD. in R Yokota, M Weiland, S Alam & J Shalf (eds), High Performance Computing - ISC High Performance 2018 International Workshops, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11203 LNCS, Springer Verlag, pp. 472-486, International Conference on High Performance Computing, ISC High Performance 2018, Frankfurt, Germany, 06/28/18. https://doi.org/10.1007/978-3-030-02465-9_34

Lessons Learned from Optimizing Kernels for Adaptive Aggregation Multi-grid Solvers in Lattice QCD. / Joó, Bálint; Kurth, Thorsten.
High Performance Computing - ISC High Performance 2018 International Workshops, Revised Selected Papers. ed. / Rio Yokota; Michèle Weiland; Sadaf Alam; John Shalf. Springer Verlag, 2018. p. 472-486 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11203 LNCS).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Lessons Learned from Optimizing Kernels for Adaptive Aggregation Multi-grid Solvers in Lattice QCD

AU - Joó, Bálint

AU - Kurth, Thorsten

PY - 2018

Y1 - 2018

N2 - In recent years, adaptive aggregation multi-grid (AAMG) methods have become the gold standard for solving the Dirac equation in Lattice QCD (LQCD) using Wilson-Clover fermions. These methods are able to overcome the critical slowing down as quark masses approach their physical values and are thus the go-to method for performing Lattice QCD calculations at realistic physical parameters. In this paper we discuss the optimization of a specific building block for implementing AAMG for Wilson-Clover fermions from LQCD, known as the coarse restrictor operator, on contemporary Intel processors featuring large SIMD widths and high thread counts. We will discuss in detail the efficient use of OpenMP and Intel vector intrinsics in our attempts to exploit fine grained parallelism on the coarsest levels. We present performance optimizations and discuss the ramifications for implementing a full AAMG stack on Intel Xeon Phi Knights Landing and Skylake processors.

AB - In recent years, adaptive aggregation multi-grid (AAMG) methods have become the gold standard for solving the Dirac equation in Lattice QCD (LQCD) using Wilson-Clover fermions. These methods are able to overcome the critical slowing down as quark masses approach their physical values and are thus the go-to method for performing Lattice QCD calculations at realistic physical parameters. In this paper we discuss the optimization of a specific building block for implementing AAMG for Wilson-Clover fermions from LQCD, known as the coarse restrictor operator, on contemporary Intel processors featuring large SIMD widths and high thread counts. We will discuss in detail the efficient use of OpenMP and Intel vector intrinsics in our attempts to exploit fine grained parallelism on the coarsest levels. We present performance optimizations and discuss the ramifications for implementing a full AAMG stack on Intel Xeon Phi Knights Landing and Skylake processors.

UR - http://www.scopus.com/inward/record.url?scp=85066118596&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-02465-9_34

DO - 10.1007/978-3-030-02465-9_34

M3 - Conference contribution

AN - SCOPUS:85066118596

SN - 9783030024642

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 472

EP - 486

BT - High Performance Computing - ISC High Performance 2018 International Workshops, Revised Selected Papers

A2 - Yokota, Rio

A2 - Weiland, Michèle

A2 - Alam, Sadaf

A2 - Shalf, John

PB - Springer Verlag

T2 - International Conference on High Performance Computing, ISC High Performance 2018

Y2 - 28 June 2018 through 28 June 2018

ER -

Joó B, Kurth T. Lessons Learned from Optimizing Kernels for Adaptive Aggregation Multi-grid Solvers in Lattice QCD. In Yokota R, Weiland M, Alam S, Shalf J, editors, High Performance Computing - ISC High Performance 2018 International Workshops, Revised Selected Papers. Springer Verlag. 2018. p. 472-486. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-02465-9_34

Lessons Learned from Optimizing Kernels for Adaptive Aggregation Multi-grid Solvers in Lattice QCD

Abstract

Publication series

Conference

Funding

Access to Document

Other files and links

Fingerprint

Cite this