Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Idomura, Yasuhiro
no journal, ,
Under the Post-K project, a Gyrokinetic Toroidal 5D full-f Eulerian code GT5D has been developed towards exascale burning plasma simulations on the Post-K machine. In this talk, we review the present status on new computational techniques on GT5D, in which Krylov based sparse matrix solvers for a semi-implicit time integration method occupy the most of computational cost. Recently, we introduced communication-avoiding (CA) Krylov methods, and improved the scalability of GT5D on the latest many-core platforms. We also ported the CA-Krylov solver on the latest GPU platforms, and achieved speedup by more than an order of magnitude.
Matsuoka, Seikichi*; Idomura, Yasuhiro; Satake, Shinsuke*; Honda, Mitsuru*; Suzuki, Yasuhiro*
no journal, ,
A Gyrokinetic Toroidal 5D full-f Eulerian code GT5D is extended for three dimensional magnetic configurations such as stellarators and tokamaks with perturbed magnetic fields. An interface program for three dimensional equilibrium magnetic configuration code VMEC is constructed to generate a vector potential used in GT5D. The developed code is verified through benchmark calculations of neoclassical transport and geodesic acoustic modes in a Large Helical Device configuration. The results show quantitative agreements against the three dimensional delta-f particle code FORTEC3D.
Imamura, Toshiyuki*; Idomura, Yasuhiro; Ina, Takuya*; Yamashita, Susumu; Onodera, Naoyuki; Ali, Y.*; Yamada, Susumu
no journal, ,
Towards exascale computing on the Post-K computer, a novel matrix solvers are developed by using communication avoiding algorithms. In this talk, we review two main approaches used in the three-dimensional thermal hydraulic multi-phase CFD code, JUPITER. One is a communication avoiding Krylov sub-space method, in which multiple basis vectors are generated and orthogonalized at once to reduce global collective communications. The other is a Krylov sub-space method with multi-grid preconditioning, which dramatically improve convergence property and reduce the number of iterations, and thus, global collective communications. We compare these approaches on the latest many core platform.