Scalability Study on Large-Scale Parallel Finite Element Computing in PANDA Frame

Abstract:

A Finite-element parallel computing frame—PANDA and its implementation processes are introduced. To validate the parallel performance of the PANDA frame, a series of tests were carried out to obtain the computing scale and the speedup ratios. First, three different large-scale freedom degree models (i.e. 1.83 million, 7 million and 10 million) of a typical engineering clamp were created in MSC.Patran and were translated into geometric-grid files that can be identified in PANDA frame. Second, Linear static parallel computations of the three cases were successfully carried out on large parallel computers with preconditioned conjugate gradient methods in PANDA frame. The speedup ratios of the three cases were obtained with a maximum process number of 64. The results show that the PANDA frame is competent for carrying out large-scale parallel computing of 10 million freedom degrees. In each scale，the parallel computing is nearly linearly accelerated along with the increase of process numbers, moreover, a super-linear speedup appears in some cases. The speedup curves show that the linear degree increases when the computing scale enlarges. The influence of different communication bandwidths on computing efficiency was also discussed. All the testing results indicate that the PANDA frame has excellent parallel performance and favorable computing scalability.

• Periodical: Applied Mechanics and Materials (Volumes 117-119)
• Pages: 489-492
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.117-119.489
• Online since: October 2011
• Authors: Xuan Hua Fan, Rui An Wu, Pu Chen
• Keywords: Finite Element, Large-Scale Parallel Computing, Preconditioned Conjugate Gradient Method, Scalability, Speedup Ratio

References

[1] R.Q. Zhang. Computational Structural Mechanics and Applications, Vol. 4 (1995) No. 12, p.477.(in Chinese)

[2] J.W. Manke. Parallel Computing, Vol.27 (2001), p.329.

[3] O. Yasar, H. Dag. Parallel Computing, Vol.33 (2007), p.81.

[4] H. Carter, Edwards. SIERRA Framework Version 3: Core Services Theory and Design, SAND 2002-3616.

[5] Tahoe home page on http://tahoe.ca.sandia.gov.

[6] G. M. Shi, R. Wu and K. Y. Wang: 9th World Congress on Computational Mechanics and 4th Asian Pacific Congress on Computational Mechanics(Sydney, Australia, July 19–23, 2010), IOP Publishing.

DOI: 10.1088/1757-899X/10/1/012007

[7] PHG home page on ftp://159.226.92.111/pub/RPMS

[8] X.H. Fan, R.A. Wu and Z.M. Hao. Chinese Journal of Computation Physics, Vol. 26 (2009) No.5, p.699. (in Chinese)