Enabling Large Scale Simulations for Particle Accelerators
نویسندگان
چکیده
International high-energy particle physics research centers, like CERN and Fermilab, require excessive studies simulations to plan for the upcoming upgrades of world's largest accelerators, design future machines given technological challenges tight budgetary constraints. The Beam Longitudinal Dynamics (BLonD) simulator suite incorporates most detailed complex phenomena in field longitudinal beam dynamics, required providing extremely accurate predictions. Modern dynamics dictate longer, larger numerous simulation draw meaningful conclusions that will drive baseline choices daily operation current projects. These are time consuming, would be impractical perform without a High-Performance Computing oriented framework. In this article, at first, we evaluate highly-optimized distributed version BLonD. We combine approximate computing techniques, leverage dynamic load-balancing scheme relax synchronization improve scalability. addition, employ GPUs accelerate implementation. highly optimized supercomputing system demonstrate speedups more than two orders magnitude when run on 32 GPU platforms, w.r.t. previous state-of-art. By driving wide range new studies, proposed high performance forms an invaluable tool accelerator physicists.
منابع مشابه
Enabling Very-Large Scale Earthquake Simulations on Parallel Machines
The Southern California Earthquake Center initiated a major largescale earthquake simulation called TeraShake. The simulations propagated seismic waves across a domain of 600x300x80 km at 200 meter resolution, some of the largest and most detailed earthquake simulations of the southern San Andreas fault. The output from a single simulation may be as large as 47 terabytes of data and 400,000 fil...
متن کاملEnabling Large-scale Network Simulations: A Selective Abstraction Approach
ions Abstraction is possible because most protocol architectures are layered such that one protocol makes use of another. In design or evaluation of a level-n protocol, we need information provided by level n-1 and below, but not necessarily (depending on the research questions) all the details of the lower level protocols. For instance, a multicast transport protocol may need multicast routing...
متن کاملLarge Scale Traffic Simulations
SFI WORKING PAPER: 1996-10-080 SFI Working Papers contain accounts of scientific work of the author(s) and do not necessarily represent the views of the Santa Fe Institute. We accept papers intended for publication in peer-reviewed journals or proceedings volumes, but not papers that have already appeared in print. Except for papers by our external faculty, papers must be based on work done at ...
متن کاملLarge Scale Simulations
In this chapter we will focus on molecular dynamics (MD) simulations with large numbers of atoms (N > 1000). Typically, ab initio or First Principles Molecular Dynamics (FPMD) calculations are performed with smaller clusters of say 100 or fewer atoms. This choice is largely due to limitations in computing resources. The simulation procedure we will explore in this chapter is known as Empirical ...
متن کاملSolution accelerators for large-scale three-dimensional electromagnetic inverse problems
We provide a framework for preconditioning nonlinear three-dimensional electromagnetic inverse scattering problems using nonlinear conjugate gradient (NLCG) and limited memory (LM) quasi-Newton methods. Key to our approach is the use of an approximate adjoint method that allows for an economical approximation of the Hessian that is updated at each inversion iteration. Using this approximate Hes...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: IEEE Transactions on Parallel and Distributed Systems
سال: 2022
ISSN: ['1045-9219', '1558-2183', '2161-9883']
DOI: https://doi.org/10.1109/tpds.2022.3192707