DOI: https://doi.org/10.26089/NumMet.v18r434

Scalability study of FlowVision on the cluster with Angara interconnect

Authors

  • V.S. Akimov
  • D.P. Silaev
  • A.S. Simonov
  • A.S. Semenov

Keywords:

scalability
Flow Vision
CFD
gas dynamics
cluster
supercomputer
interconnect
Angara

Abstract

The scalability of computations in FlowVision CFD software on the Angara-C1 cluster equipped with Angara interconnect is studied. Several test problems with 260 thousand, 5.5 million and 26.8 million computational cells are considered. Computations in FlowVision are performed using a new solver of linear systems based on the algebraic multigrid (AMG) method. It is shown that the special FlowVision’s technology named «Dynamic balancing» significantly improves the performance of computations if the peculiarities of the problem promote the non-uniform loading of CPUs. The Angara-C1 cluster demonstrates the excellent performance and scalability characteristics comparable with its analogues based on the 4х FDR Infiniband interconnect.

New computational technologies

Downloads

Published

2017-10-07

Issue

Vol. 18 (2017): Issue 4.

Section

Section 1. Numerical methods and applications

Author Biographies

V.S. Akimov

OOO Vychislitelnaya inzhenernaya platforma
• Engineer

D.P. Silaev

OOO Vychislitelnaya inzhenernaya platforma
• Lead Developer

A.S. Simonov

Research Center for Electronic Computing
• First Deputy CEO

A.S. Semenov

Research Center for Electronic Computing
• Deputy Department Head

References

  1. S. Derradji, T. Palfer-Sollier, J.-P. Panziera, et al., “The BXI Interconnect Architecture,” in Proc. IEEE 23rd Annual Symp. on High-Performance Interconnects, Santa Clara, USA, August 26-28, 2015 (IEEE Press, Washington, DC, 2015),
    doi 10.1109/HOTI.2015.15
  2. H. Fröning, M. Nussle, H. Litz, et al., “On Achieving High Message Rates,” in Proc. 13th IEEE/ACM Int. Symp. on Cluster, Cloud, and Grid Computing, Delft, The Netherlands, May 13-16, 2013 (IEEE, New York, 2013),
    doi 10.1109/CCGrid.2013.43
  3. I. A. Zhabin, D. V. Makagon, D. A. Polyakov, et al., “First Generation of Angara High-Speed Interconnection Network,” Naukoemkie Tekhnol., No. 1, 21-27 (2014).
  4. A. I. Slutskin, A. S. Simonov, I. A. Zhabin, et al., “Development of the ES8430 Angara Interconnect for Future Russian Supercomputers,” Usp. Sovr. Radioelektron., No. 1, 6-10 (2012).
  5. A. A. Agarkov, T. F. Ismagilov, D. V. Makagon, et al., “Performance Evaluation of the Angara Interconnect,” in Proc. Int. Conf. on Russian Supercomputing Days, Moscow, Russia, September 26-27, 2016 (Mosk. Gos. Univ., Moscow, 2016), pp. 626-639.
  6. FlowVision Help.
    https://flowvision.ru/webhelp/fvru_30905 . Cited November 14, 2017.
  7. G. B. Sushko and S. A. Kharchenko, “Experimental Investigation of MPI+threads Implementation of the Algorithm for Solving Systems of Linear Equations Occurring in FlowVision when Solving CFD Problems Using the Chebyshev Supercomputer of Moscow University,” in Proc. Int. Conf. on Parallel Computational Technologies, Nizhnii Novgorod, Russia, March 30-April 3, 2009 (South Ural State Univ., Chelyabinsk, 2009), pp. 316-324.
  8. M. J. C. Henshaw, “M219 Cavity Case: Verification and Validation Data for Computational Unsteady Aerodynamics,” Tech. Rep. RTO-TR-26, AC/323(AVT)TP/19, QinetiQ, UK, 453-472 (2002).

License

Copyright (c) 2017 Вычислительные методы и программирование

Creative Commons License

This work is licensed under a Creative Commons Attribution 4.0 International License.