Paper  Title  Page 

SUPAF11 
Computer Architecture Independent Adaptive Geometric Multigrid Solver for AMRPIC  


Funding: SNSF project 200021_{1}59936 The accurate and efficient simulation of neighboring bunch effects in high intensity cyclotrons requires to solve largescale Nbody problems of O(10^{9}…10^{z}EhNZeHn) particles coupled with Maxwell’s equations. In order to capture the effects of halo creation and evolution of such simulations with standard particleincell models an extremely fine mesh with O(10^{8}…10^{9}) grid points is necessary to meet the condition of high resolution. This requirement represents a waste of memory in regions of void, therefore, the usage of blockstructured adaptive mesh refinement algorithms is more suitable. The Nbody problem is then solved on a hierarchy of levels and grids using geometric multigrid algorithms. We show benchmarks of a new implementation of an adaptive geometric multigrid algorithm using 2nd generation Trilinos packages that ran on Piz Daint with O(10^{4}…10^{5}) cores. 

Slides SUPAF11 [6.094 MB]  
Export •  reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
SUPAF12 
Surrogate Models for Beam Dynamics in Charged Particle Accelerators  


Highfidelity, PICbased beam dynamics simulations are time and resource intensive. Consider a high dimensional search space, that is far too large to probe with such a high resolution simulation model. We demonstrate that a coarse sampling of the search space can produce surrogate models, which are accurate and fast to evaluate. In constructing the surrogate models, we use artificial neural networks [1] and multivariate polynomial chaos expansion [2]. The performance of both methods are demonstrated in a comparison with highfidelity simulations, using OPAL, of the Argonne Wakefield Accelerator [3]. We claim that such surrogate models are good candidates for accurate online modeling of large, complex accelerator systems. We also address how to estimate the accuracy of the surrogate model and how to refine the surrogate model under changing machine conditions. [1] A. L. Edelen et al., arXiv:1610.06151[physics.acc ph] [2] A. Adelmann, arXiv:1509.08130v6[physics.acc ph] [3] N. Neveu et al., 2017 J. Phys.: Conf. Ser. 874 012062  
Slides SUPAF12 [11.505 MB]  
Export •  reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
SUPLG01  Computational Accelerator Physics: On the Road to Exascale  113 


The first conference in what would become the ICAP series was held in 1988. At that time the most powerful computer in the world was a Cray YMP with 8 processors and a peak performance of 2 gigaflops. Today the fastest computer in the world has more than 2 million cores and a theoretical peak performance of nearly 200 petaflops. Compared to 1988, performance has increased by a factor of 100 million, accompanied by huge advances in memory, networking, big data management and analytics. By the time of the next ICAP in 2021 we will be at the dawn of the Exascale era. In this talk I will describe the advances in Computational Accelerator Physics that brought us to this point and describe what to expect in regard to High Performance Computing in the future. This writeup as based on my presentation at ICAP’18 along with some additional comments that I did not include originally due to time constraints.  
Slides SUPLG01 [25.438 MB]  
DOI •  reference for this paper ※ https://doi.org/10.18429/JACoWICAP2018SUPLG01  
About •  paper received ※ 14 November 2018 paper accepted ※ 07 December 2018 issue date ※ 26 January 2019  
Export •  reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  
TUPAG06 
Parallel Algorithms for Solving Nonlinear Eigenvalue Problems in Accelerator Cavity Simulations  


We present an efficient and reliable algorithm for solving a class of nonlinear eigenvalue problems arising from the modeling of particle accelerator cavities. The eigenvalue nonlinearity in these problems results from the use of waveguides to couple external power sources or to allow certain excited electromagnetic modes to exit the cavity. We use a rational approximation to reduce the nonlinear eigenvalue problem first to a rational eigenvalue problem. We then apply a special linearization procedure to turn the rational eigenvalue problem into a larger linear eigenvalue problem with the same eigenvalues, which can be solved by existing iterative methods. By using a compact scheme to represent both the linearized operator and the eigenvectors to be computed, we obtain a numerical method that only involves solving linear systems of equations of the same dimension as the original nonlinear eigenvalue problem. We refer to this method as a compact rational Krylov (CORK) method. We implemented the CORK method in the Omega3P module of the Advanced Computational Electromagnetic 3D Parallel (ACE3P) simulation suite and validated it by comparing the computed cavity resonant frequencies and damping Q factors of a small model problem to those obtained from a fitting procedure that uses frequency responses computed by another ACE3P module called S3P. We also used the CORK method to compute trapped modes damped in an ideal eight 9cell SRF cavity cryomodule. This was the first time it was possible to compute these modes directly. The damping Q factors of the computed modes match well with those measured in experiments and the difference in resonant frequencies is within the range introduced by cavity imperfection. Therefore, the CORK method is an extremely valuable tool for computational cavity design.  
Slides TUPAG06 [1.252 MB]  
Export •  reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)  