Search

Dispersion-free ultra-high order FFT-based Maxwell solvers have recently proven to be paramount to a large range of applications, including the high-fidelity modeling of high-intensity laser–matter interactions with Particle-In-Cell (PIC) codes. To enable a massively parallel scaling of these solvers, a novel parallelization technique was recently proposed, which consists in splitting the simulation domain into several processor sub-domains, with guard regions appended at each sub-domain boundary. Maxwell's equations are advanced independently on each sub-domain using local shared-memory FFTs (instead of a single distributed global FFT). This implies small truncation errors at sub-domain boundaries, the amplitude of which depends on guard regions sizes and order of the Maxwell solver. For moderate guard region sizes, this 'local' technique proved to be highly scalable on up to a million cores and notably enabled the 3D modelingof so-called plasma mirrors, for which 8 guard cells only were enough to prevent truncation error growth. Yet, for other applications, the required number of guard cells might be much higher, which would severely limit the parallel efficiency of this technique due to the large volume of guard cells to be exchanged between sub-domains. In this context, we propose a novel parallelization technique that ensures very good scaling of FFT-based solvers with an arbitrarily high number of guard cells. Our 'hybrid' technique consists in performing distributed FFTs on local groups of processors with guard regions now appended to boundaries of each group of processors. It uses a dual domain decomposition method for the Maxwell solver and other parts of the PIC cycle to keep the simulation load-balanced. This 'hybrid' technique was implemented in the open source exascale library PICSAR. Benchmarks show that for a large number of guard cells (> 16), the 'hybrid' technique offers up to × 3 speed-up and × 8 memory savings compared to the 'local' one. [ABSTRACT FROM AUTHOR]

This paper presents XGCm, a new unstructured mesh gyrokinetic Particle-in-Cell (PIC) code for modeling fusion plasma. The physical models and aspects of the numerical methods employed are the same as those used in the X-point gyrokinetic code, XGC. The core difference is that XGCm builds on an unstructured mesh-centric infrastructure that supports a distributed mesh, making it scalable in both the number of mesh elements and number of particles. A second advantage of an unstructured mesh infrastructure is its performance is not degraded when generally graded or anisotropic meshes are used. The switch from a particle-centric data infrastructure to a distributed mesh-centric infrastructure required the introduction of new methods to execute core PIC operations and substantial modifications to a number of key algorithms from those used in XGC. First, we present the methods and algorithms used in the development of XGCm, which performs all key computing steps on the GPU accelerators. GPU accelerators are providing the main computational power of current generation U.S. Department of Energy supercomputers. Secondly, we perform code validation and test using the circular geometry cyclone base case and realistic DIII-D geometry case, respectively. The turbulence growth rate shows excellent agreement with existing XGC result in the first case, while ion temperature gradient turbulence growth is further demonstrated in the second case. Finally, we present weak scaling results, using up to full system (27,648 GPUs) of the Oak Ridge National Laboratory's Summit supercomputer. [ABSTRACT FROM AUTHOR]

We present a new hybrid algorithm implemented in the code CHIEF ( C ode H ybrid with I nertial E lectron F luid) for simulations of electron–ion plasmas. The algorithm treats the ions kinetically, modeled by the Particle-in-Cell (PiC) method, and electrons as an inertial fluid, modeled by electron fluid equations without any of the approximations used in most of the other hybrid codes with an inertial electron fluid. This kind of code is appropriate to model a large variety of quasineutral plasma phenomena where the electron inertia and/or ion kinetic effects are relevant. We present here the governing equations of the model, how these are discretized and implemented numerically, as well as six test problems to validate our numerical approach. Our chosen test problems, where the electron inertia and ion kinetic effects play the essential role, are: 0) Excitation of parallel eigenmodes to check numerical convergence and stability, 1) parallel (to a background magnetic field) propagating electromagnetic waves, 2) perpendicular propagating electrostatic waves (ion Bernstein modes), 3) ion beam right-hand instability (resonant and non-resonant), 4) ion Landau damping, 5) ion firehose instability, and 6) 2D oblique ion firehose instability. Our results reproduce successfully the predictions of linear and non-linear theory for all these problems, validating our code. All properties of this hybrid code make it ideal to study multi-scale phenomena between electron and ion scales such as collisionless shocks, magnetic reconnection and kinetic plasma turbulence in the dissipation range above the electron scales. [ABSTRACT FROM AUTHOR]

The microscopic particle-in-cell (MicPIC) method was developed to model classical light–matter interaction in strongly-coupled plasma systems. It effectively overcomes the limitations of the particle-in-cell and molecular dynamics techniques by combining them into a single, unified framework to solve for both electromagnetic wave propagation and atomic-scale collision processes in a self-consistent treatment. Its effective time complexity is O ( N ) , where N is the number of model particles, which is ideal for studying the dynamics of large ensembles. In this paper, we show that through massively parallel, distributed computations, current implementations of the MicPIC approach can handle up to 10 11 particles on an IBM Blue Gene/Q computer with 65 536 physical cores. This allows modelling volumes of matter of approximately 1 μ m 3 at solid gold density, opening a wealth of potential applications of MicPIC in nanophotonics, diffractive X-ray imaging, and strong-field science. [ABSTRACT FROM AUTHOR]

In current computer architectures, data movement (from die to network) is by far the most energy consuming part of an algorithm (≈20 pJ/word on-die to ≈10,000 pJ/word on the network). To increase memory locality at the hardware level and reduce energy consumption related to data movement, future exascale computers tend to use many-core processors on each compute nodes that will have a reduced clock speed to allow for efficient cooling. To compensate for frequency decrease, machine vendors are making use of long SIMD instruction registers that are able to process multiple data with one arithmetic operator in one clock cycle. SIMD register length is expected to double every four years. As a consequence, Particle-In-Cell (PIC) codes will have to achieve good vectorization to fully take advantage of these upcoming architectures. In this paper, we present a new algorithm that allows for efficient and portable SIMD vectorization of current/charge deposition routines that are, along with the field gathering routines, among the most time consuming parts of the PIC algorithm. Our new algorithm uses a particular data structure that takes into account memory alignment constraints and avoids gather/scatter instructions that can significantly affect vectorization performances on current CPUs. The new algorithm was successfully implemented in the 3D skeleton PIC code PICSAR and tested on Haswell Xeon processors (AVX2-256 bits wide data registers). Results show a factor of×2 to×2.5 speed-up in double precision for particle shape factor of orders 1-3. The new algorithm can be applied as is on future KNL (Knights Landing) architectures that will include AVX-512 instruction sets with 512 bits register lengths (8 doubles/16 singles). [ABSTRACT FROM AUTHOR]

Abstract: In the diesel particulate filters technology a key aspect is represented by the properties of the particulate matter that is collected inside their structure. The work presented is focused on the development of an innovative mathematical tool based on the particle-in-cell method (PIC) for the simulation of the soot distribution inside a single channel of a diesel particulate filter. The basic fluid dynamic equations are solved for the gas phase inside the channel using a novel technique based on the solution of the same set of equations everywhere in the system including the porous medium. This approach is presented as alternative to the more conventional methods of matching conditions across the boundary of the porous region where a Darcy-like flow is developed. The motion of the soot solid particles is instead described through a particle-by-particle approach based on Newton''s equations of motion. The coupling between the dynamics of the gas and that of the soot particles, i.e. between these two sub-models, is performed through the implementation of the particle-in-cell technique. This model allows the detailed simulation of the deposition and compaction of the soot inside the filter channels and its characterization in terms of density, permeability and thickness. The model then represents a unique tool for the optimization of the design of diesel particulate filters. The details of the technique implementation and some paradigmatic examples will be shown. [Copyright &y& Elsevier]

In this paper, we describe an efficient, massively parallel GPU implementation strategy for speeding up one-dimensional electrostatic plasma simulations based on the Particle-in-Cell method with Monte-Carlo collisions. Relying on the Roofline performance model, we identify performance-critical points of the program and provide optimised solutions. We use four benchmark cases to verify the correctness of the CUDA and OpenCL implementations and analyse their performance properties on a number of NVIDIA and AMD cards. Plasma parameters computed with both GPU implementations differ not more than 2% from each other and respective literature reference data. Our final implementations reach over 2.6 Tflop/s sustained performance on a single card, and show speed up factors of up to 200 (when using 10 million particles). We demonstrate that GPUs can be very efficiently used for simulating collisional plasmas and argue that their further use will enable performing more accurate simulations in shorter time, increase research productivity and help in advancing the science of plasma simulation. [ABSTRACT FROM AUTHOR]