The Gyrokinetic Particle Simulation of Fusion Plasmas on Tianhe-2 Supercomputer

We present novel optimizations of the fusion plasmas simulation code, GTC on Tianhe-2 supercomputer. The simulation exhibits excellent weak scalability up to 3072 31S1P Xeon Phi co-processors. An unprecedented up to 5.8× performance improvement is achieved for the GTC on Tianhe-2. An efficient particle exchanging algorithm is developed that simplifies the original iterative scheme to a direct implementation, which leads to a 7.9× performance improvement in terms of MPI communications on 1024 nodes of Tianhe-2. A customized particle sorting algorithm is presented that delivers a 2.0× performance improvement on the co-processor for the kernel relating to the particle computing. A smart offload algorithm that minimizes the data exchange between host and co-processor is introduced. Other optimizations like the loop fusion and vectorization are also presented.

[1]  Weixing Wang,et al.  NERSC's Impact on Advances of Global Gyrokinetic PIC Codes for Fusion Energy Research , 2015, Computing in Science & Engineering.

[2]  Z. Lin,et al.  Size scaling of turbulent transport in magnetically confined plasmas. , 2002, Physical review letters.

[3]  S. Ethier,et al.  Gyrokinetic particle-in-cell simulations of plasma microturbulence on advanced computing platforms , 2005 .

[4]  Samuel Williams,et al.  Gyrokinetic toroidal simulations on leading multi- and manycore HPC systems , 2011, 2011 International Conference for High Performance Computing, Networking, Storage and Analysis (SC).

[5]  Samuel Williams,et al.  Gyrokinetic particle-in-cell optimization on emerging multi- and manycore platforms , 2011, Parallel Comput..

[6]  Samuel Williams,et al.  Analysis and optimization of gyrokinetic toroidal simulations on homogenous and heterogenous platforms , 2013, Int. J. High Perform. Comput. Appl..

[7]  Bei Wang,et al.  Scientific Discovery in Fusion Plasma Turbulence Simulations at Extreme Scale , 2014, Computing in Science & Engineering.

[8]  S. Ethier,et al.  Verification of Gyrokinetic Particle Simulation of Device Size Scaling of Turbulent Transport , 2012 .

[9]  Leonid Oliker,et al.  Petascale Parallelization of the Gyrokinetic Toroidal Code , 2010, HiPC 2010.

[10]  T. Hahm,et al.  Turbulent transport reduction by zonal flows: massively parallel simulations , 1998, Science.

[11]  Samuel Williams,et al.  Kinetic turbulence simulations at extreme scale on leadership-class systems , 2013, 2013 SC - International Conference for High Performance Computing, Networking, Storage and Analysis (SC).

[12]  Zhihong Lin,et al.  Turbulent transport of trapped-electron modes in collisionless plasmas. , 2009, Physical review letters.

[13]  Zhihong Lin,et al.  Porting the 3D gyrokinetic particle-in-cell code GTC to the NEC SX-6 vector architecture: perspectives and challenges , 2004, Comput. Phys. Commun..

[14]  John Mellor-Crummey,et al.  Managing locality in grand challenge applications: a case study of the gyrokinetic toroidal code , 2008 .

[15]  Kevin J. Bowers,et al.  Accelerating a paricle -in-cell simulation using a hybrid counting sort , 2001 .

[16]  Ihor Holod,et al.  Radial localization of toroidicity-induced Alfvén eigenmodes. , 2013, Physical review letters.

[17]  Leonid Oliker,et al.  Large-scale gyrokinetic particle simulation of microturbulence in magnetically confined fusion plasmas , 2008, IBM J. Res. Dev..

[18]  Xin Liu,et al.  Heterogeneous Programming and Optimization of Gyrokinetic Toroidal Code and Large-Scale Performance Test on TH-1A , 2013, ISC.

[19]  R. Aymar,et al.  The ITER project , 1997 .

[20]  Z. Lin,et al.  Wave-particle decorrelation and transport of anisotropic turbulence in collisionless plasmas. , 2007, Physical review letters.

[21]  Zhihong Lin,et al.  Transport of energetic particles by microturbulence in magnetized plasmas. , 2008, Physical review letters.

[22]  Z. Lin,et al.  Nonlinear frequency oscillation of Alfvén eigenmodes in fusion plasmas. , 2012, Physical review letters.