It is now widely believed that low frequency turbulence developing from smallscale instabilities is responsible for the phenomenon of anomalous transport generally observed in magnetic confinement fusion experiments.The microinstabilities are driven by gradients of equilibrium density, ion and electron temperatures and magnetic field strength.Gyrokinetic theory is based on the Vlasov-Maxwell equations and, consistent with the ordering, averages out the fast particle gyromotion, reducing the phase space from 6 to 5 dimensions.Solving the resulting equations is a non-trivial task.Difficulties are associated with the magnetic confinement geometry, the strong disparities in space and time scales perpendicular and parallel to B, the different time scales of ion and electron dynamics, and the complex nonlinear behaviour of the system.The main numerical methods are briefly presented together with some recent developments and improvements to the basic algorithms.Recent results are shown, with emphasis on the roles of zonal E ×B flows, of parallel nonlinearity and of toroidal coupling on the saturation of ion temperature gradient (ITG) driven turbulence in tokamaks.
The left-preconditioned communication avoiding conjugate gradient (LP-CA-CG) method is applied to the pressure Poisson equation in the multiphase CFD code JUPITER. The arithmetic intensity of the LP-CA-CG method is analyzed, and is dramatically improved by loop splitting for inner product operations and for three term recurrence operations. Two LPCA-CG solvers with block Jacobi preconditioning and with underlap preconditioning are developed. The former is developed based on a hybrid CA approach, in which CA is applied only to global collective communications for inner product operations. The latter is a full CA approach, in which CA is applied also to local point-to-point communications in sparse matrix-vector (SpMV) operations and preconditioning. CA-SpMV requires additional computation for overlapping regions. CA-preconditiong is enabled by underlap preconditioning, which approximates preconditioning for overlapping regions by point Jacobi preconditioning. It is shown that on the K computer, the former is faster, because the performance of local point-to-point communications scales well, and the convergence property becomes worse with underlap preconditioning. The LP-CA-CG solver shows good strong scaling up to 30,000 nodes, where the LP-CA-CG solver achieved higher performance than the original CG solver by reducing the cost of global collective communications by 69 percent.
This dataset contains the two dimensional fluid simulation data for AMR-Net. The inputs of simulations are signed distance functions (SDFs) and the outputs are 2D flow fields u and v. In each simulation, 1-5 objects are placed randomly at the center of the computational domain. The objects are randomly chosen from circles, ellipses, rectangles and rounded rectangles. Each hdf5 file includes the pair of input and output data. The details of the dataset is described in the Github page.
The influence of plasma size on global ion temperature gradient turbulence is studied with the full- f Eulerian code GT5D (Idomura et al 2009 Nucl. Fusion 49 065029 ). The gyrokinetic model includes a consistent neoclassical electric field as well as a fixed-power source operator, enabling long-time simulations with self-consistent turbulent transport and equilibrium profiles. The effects of plasma size (from ρ * = 1/100 to ρ * = 1/225) are studied by scaling the minor radius a and the input power. For the first time, worse-than-Bohm scaling is observed under experimentally realistic conditions. For all plasma sizes, avalanches propagate over significant radii but their propagation depends on the radial electric shear. It is found that this quantity does not scale with ρ * due to the building up of intrinsic momentum. Such a dependence can be inferred from a force balance relation, which remains approximately valid in nonlinear simulations. An adaptive parallel momentum source has been implemented in GT5D to damp the parallel momentum profile. The new scan then reveals that the radial electric shear scales with ρ * while the transport is globally higher. These simulations therefore suggest that intrinsic momentum reduces heat transport. This work also addresses another important issue in gyrokinetics: it is shown that for fixed initial physical parameters the turbulent quasi-steady-state is statistically independent of the initial conditions.
GT5D is a nuclear fusion simulation program which aims to analyze the turbulence phenomena in tokamak plasma. In this research, we optimize it for GPU clusters with multiple GPUs on a node. Based on the profile result of GT5D on a CPU node, we decide to offload the whole of the time development part of the program to GPUs except MPI communication. We achieved 3.37 times faster performance in maximum in function level evaluation, and 2.03 times faster performance in total than the case of CPU-only execution, both in the measurement on high density GPU cluster HA-PACS where each computation node consists of four NVIDIA M2090 GPUs and two Intel Xeon E5-2670 (Sandy Bridge) to provide 16 cores in total. These performance improvements on single GPU corresponds to four CPU cores, not compared with a single CPU core. It includes 53% performance gain with overlapping the communication between MPI processes with GPU calculation.
'%3e%3cpath fill='%23012169' d='M0 0v30h60V0z'/%3e%3cpath stroke='%23fff' stroke-width='6' d='m0 0 60 30m0-30L0 30'/%3e%3cpath stroke='%23C8102E' stroke-width='4' d='m0 0 60 30m0-30L0 30' clip-path='url(%23b)'/%3e%3cpath stroke='%23fff' stroke-width='10' d='M30 0v30M0 15h60'/%3e%3cpath stroke='%23C8102E' stroke-width='6' d='M30 0v30M0 15h60'/%3e%3c/g%3e%3c/svg%3e)
