Plasma and Fusion Research
Volume 8, 1403150 (2013)
Regular Articles
- Japan Atomic Energy Agency, Rokkasho 039-3212, Japan
- 1)
- National Institute for Fusion Science, Toki 509-5292, Japan
Abstract
One of the important phenomena in magnetically-confined fusion plasma is plasma turbulence, which causes particle and heat transport and degrades plasma confinement. To address multi-scale turbulence including temporal and spatial scales of electrons and ions, we extend our gyrokinetic Vlasov simulation code GKV to run efficiently on peta-scale supercomputers. A key numerical technique is the parallel Fast Fourier Transform (FFT) required for parallel spectral calculations, where masking of the cost of inter-node transpose communications is essential to improve strong scaling. To mask communication costs, computation-communication overlap techniques are applied for FFTs and transpose with the help of the hybrid parallelization of message passing interface and open multi-processing. Integrated overlaps including whole spectral calculation procedures show better scaling than simple overlaps of FFTs and transpose. The masking of communication costs significantly improves strong scaling of the GKV code, and makes substantial speed-up toward multi-scale turbulence simulations.
Keywords
computation-communication overlap, parallel fast Fourier transform, parallel spectral calculation, MPI/OpenMP hybrid parallelization, Vlasov simulation
Full Text
References
- [1] X. Garbet, Y. Idomura, L. Villard and T.-H. Watanabe, Nucl. Fusion 50, 043002 (2010).
- [2] T.-H. Watanabe and H. Sugama, Nucl. Fusion 46, 24 (2006).
- [3] D.G. Fox and S.A. Orszag, J. Comput. Phys. 11, 612 (1973).
- [4] J.W. Cooley and J.W. Tukey, Math. Comput. 19, 297 (1965).
- [5] C. Calvin, Parallel Comput. 22, 1255 (1996).
- [6] Z. Yin, L. Yuan and T. Tang, J. Comput. Phys. 210, 325 (2005).
- [7] M. Iovieno, C. Cavazzoni and D. Tordella, Comput. Phys. Commun. 141, 365 (2001).
- [8] P. Wapperom, A.N. Beris and M.A. Straka, Parallel Comput. 32, 1 (2006).
- [9] A. Danalis, K.Y. Kim, L. Pollock and M. Swany, Proc. IEEE/ACM Int. Conf. High Perform. Comput. (SC2005), Seattle, USA, pp. 58 (2005).
- [10] P.D. Mininni, D. Rosenberg, R. Reddy and A. Pouquet, Parallel Comput. 37, 316 (2011).
- [11] T. Adachi, N. Shida, K. Miura, S. Sumimoto, A. Uno, M. Kurokawa, F. Shoji and M. Yokokawa, Comput. Sci. Res. Dev. 28, 147 (2013).
- [12] Y. Idomura, M. Nakata, S. Yamada, M. Machida, T. Imamura, T.-H. Watanabe, M. Nunami, H. Inoue, S. Tsutsumi, I. Miyoshi and N. Shida, Int. J. High Perform. Comput. Appl., DOI: 10.1177/1094342013490973 (2013).
- [13] S. Maeyama, A. Ishizawa, T.-H. Watanabe, N. Nakajima, S. Tsuji-Iio and H. Tsutsui, Comput. Phys. Commun., DOI: 10.1016/j.cpc.2013.06.014 (2013).
- [14] M.A. Beer, S.C. Cowley and G.W. Hammett, Phys. Plasmas 2, 2687 (1995).
- [15] M. Frigo and S.G. Johnson, Proc. IEEE 93, 216 (2005).
- [16] T.-H. Watanabe, Y. Todo and W. Horton, Plasma Fusion Res. 3, 061 (2008).
This paper may be cited as follows:
Shinya MAEYAMA, Tomohiko WATANABE, Yasuhiro IDOMURA, Motoki NAKATA, Masanori NUNAMI and Akihiro ISHIZAWA, Plasma Fusion Res. 8, 1403150 (2013).