1. |
- Bastrakov, S., et al.
(author)
-
Performance aspects of collocated and staggered grids for particle-in-cell plasma simulation
- 2017
-
In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). - Cham : Springer International Publishing. - 1611-3349 .- 0302-9743. - 9783319629315 ; 10421, s. 94-100
-
Conference paper (peer-reviewed)abstract
- We present a computational comparison of collocated and staggered uniform grids for particle-in-cell plasma simulation. Both types of grids are widely used, and numerical properties of the corresponding solvers are well-studied. However, for large-scale simulations performance is also an important factor, which is the focus of this paper. We start with a baseline implementation, apply widely-used techniques for performance optimization and measure their efficacy for both grids on a high-end Xeon CPU and a second-generation Xeon Phi processor. For the optimized version the collocated grid outperforms the staggered one by about 1.5x on both Xeon and Xeon Phi. The speedup on the Xeon Phi processor compared to Xeon is about 1.9x.
|
|
2. |
- Gonoskov, Arkady, 1984, et al.
(author)
-
Extended particle-in-cell schemes for physics in ultrastrong laser fields : Review and developments
- 2015
-
In: Physical Review E. Statistical, Nonlinear, and Soft Matter Physics. - 1539-3755 .- 1550-2376. ; 92:2
-
Journal article (peer-reviewed)abstract
- We review common extensions of particle-in-cell (PIC) schemes which account for strong field phenomena in laser-plasma interactions. After describing the physical processes of interest and their numerical implementation, we provide solutions for several associated methodological and algorithmic problems. We propose a modified event generator that precisely models the entire spectrum of incoherent particle emission without any low-energy cutoff, and which imposes close to the weakest possible demands on the numerical time step. Based on this, we also develop an adaptive event generator that subdivides the time step for locally resolving QED events, allowing for efficient simulation of cascades. Further, we present a unified technical interface for including the processes of interest in different PIC implementations. Two PIC codes which support this interface, PICADOR and ELMIS, are also briefly reviewed.
|
|
3. |
- Larin, Anton, et al.
(author)
-
Load balancing for particle-in-cell plasma simulation on multicore systems
- 2018
-
In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). - Cham : Springer International Publishing. - 1611-3349 .- 0302-9743. ; 10777 LNCS, s. 145-155
-
Conference paper (peer-reviewed)abstract
- Particle-in-cell plasma simulation is an important area of computational physics. The particle-in-cell method naturally allows parallel processing on distributed and shared memory. In this paper we address the problem of load balancing on multicore systems. While being well-studied for many traditional applications of the method, it is a relevant problem for the emerging area of particle-in-cell simulations with account for effects of quantum electrodynamics. Such simulations typically produce highly non-uniform, and sometimes volatile, particle distributions, which could require custom load balancing schemes. In this paper we present a computational evaluation of several standard and custom load balancing schemes for the particle-in-cell method on a high-end system with 96 cores on shared memory. We use a test problem with static non-uniform particle distribution and a real problem with account for quantum electrodynamics effects, which produce dynamically changing highly non-uniform distributions of particles and workload. For these problems the custom schemes result in increase of scaling efficiency by up to 20% compared to the standard OpenMP schemes.
|
|
4. |
- Meyerov, I., et al.
(author)
-
Hybrid CPU + Xeon Phi implementation of the Particle-in-Cell method for plasma simulation
- 2016
-
In: Supercomputing Frontiers and Innovations. - : FSAEIHE South Ural State University (National Research University). - 2313-8734 .- 2409-6008. ; 3:3, s. 5-10
-
Journal article (peer-reviewed)abstract
- This paper presents experimental results of Particle-in-Cell plasma simulation on a hybrid system with CPUs and Intel Xeon Phi coprocessors. We consider simulation of two relevant laserdriven particle acceleration regimes using the Particle-in-Cell code PICADOR. On a node of a cluster with 2 CPUs and 2 Xeon Phi coprocessors the hybrid CPU + Xeon Phi configuration allows to fully utilize the computational resources of the node. It outperforms both CPU-only and Xeon Phi-only configurations with the speedups between 1.36 x and 1.68 x.
|
|
5. |
- Surmin, I., et al.
(author)
-
Co-design of a particle-in-cell plasma simulation code for Intel Xeon Phi: A first look at Knights landing
- 2016
-
In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). - Cham : Springer International Publishing. - 1611-3349 .- 0302-9743. - 9783319499550 ; 10049, s. 319-329
-
Conference paper (peer-reviewed)abstract
- Three dimensional particle-in-cell laser-plasma simulation is an important area of computational physics. Solving state-of-the-art problems requires large-scale simulation on a supercomputer using specialized codes. A growing demand in computational resources inspires research in improving efficiency and co-design for supercomputers based on many-core architectures. This paper presents first performance results of the particle-in-cell plasma simulation code PICADOR on the recently introduced Knights Landing generation of Intel Xeon Phi. A straightforward rebuilding of the code yields a 2.43 x speedup compared to the previous Knights Corner generation. Further code optimization results in an additional 1.89 x speedup. The optimization performed is beneficial not only for Knights Landing, but also for high-end CPUs and Knights Corner. The optimized version achieves 100 GFLOPS double precision performance on a Knights Landing device with the speedups of 2.35 x compared to a 14-core Haswell CPU and 3.47 x compared to a 61-core Knights Corner Xeon Phi.
|
|
6. |
- Surmin, I., et al.
(author)
-
Dynamic Load Balancing Based on Rectilinear Partitioning in Particle-in-Cell Plasma Simulation
- 2015
-
In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). - Cham : Springer International Publishing. - 1611-3349 .- 0302-9743. - 9783319219097 ; 9251, s. 107-119
-
Conference paper (peer-reviewed)abstract
- This paper considers load balancing in Particle-in-Cell plasma simulation on cluster systems. We propose a dynamic load balancing scheme based on rectilinear partitioning and discuss implementation of efficient imbalance estimation and rebalancing. We analyze the impact of load balancing on performance and accuracy. On a test plasma heating problem dynamic load balancing yields nearly 2 times speedup and better scaling. On the real-world plasma target irradiation simulation load balancing allows to mitigate particle resampling and thus improve accuracy of the simulation without increasing the runtime. Balancing-related overhead in both cases are under 1.5% of total run time.
|
|
7. |
- Surmin, I., et al.
(author)
-
Particle-in-Cell laser-plasma simulation on Xeon Phi coprocessors
- 2016
-
In: Computer Physics Communications. - : Elsevier BV. - 0010-4655. ; 202, s. 204-210
-
Journal article (peer-reviewed)abstract
- This paper concerns the development of a high-performance implementation of the Particle-in-Cell method for plasma simulation on Intel Xeon Phi coprocessors. We discuss the suitability of the method for Xeon Phi architecture and present our experience in the porting and optimization of the existing parallel Particle-in-Cell code PICADOR. Direct porting without code modification gives performance on Xeon Phi close to that of an 8-core CPU on a benchmark problem with 50 particles per cell. We demonstrate step-by-step optimization techniques, such as improving data locality, enhancing parallelization efficiency and vectorization leading to an overall 4.2 x speedup on CPU and 7.5 x on Xeon Phi compared to the baseline version. The optimized version achieves 16.9 ns per particle update on an Intel Xeon E5-2660 CPU and 9.3 ns per particle update on an Intel Xeon Phi 5110P. For a real problem of laser ion acceleration in targets with surface grating, where a large number of macroparticles per cell is required, the speedup of Xeon Phi compared to CPU is 1.6x.
|
|