A scalable algorithm for many-body dissipative particle dynamics using multiple general purpose graphic processing units
نویسندگان
چکیده
We present a novel algorithm for the many-body Dissipative Particle Dynamics (DPD) forces calculation which allows to efficiently scale DL_MESO software package on Multiple General Purpose Graphic Processing Units. Together with extension 64-bit integer arrays and addition of hard surface boundary conditions, proposed simulate very large complex mesoscale systems up 14 billion beads. The implementation takes advantages CUDA language stream features overlap exchange particle positions local densities computation short range forces. tested water drop between two plates system using tree main European supercomputers: Piz Daint, Marconi JUWELS. Results shows an improvement speedup compared naive 1.5x when 1024 GPUs.
منابع مشابه
General-purpose Computation Using Graphical Processing Units
Graphical Processing Units (GPU) are mainly developed to process intense graphical operations. This has been the purpose for many years; however recent advancements in the architecture of GPUs and how they process data have led to a breakthrough in how we are able to process information. General Purpose Computation using GPUs is a new method of processing information using a GPU and its many pr...
متن کاملNo-go theorem in many-body dissipative particle dynamics.
Many body dissipative particle dynamics (MDPD) is a particle-based simulation method in which the interaction potential is a sum of self energies depending on locally sampled density variables. This functional form gives rise to density-dependent pairwise forces; however, not all such force laws are derivable from a potential, and the integrability condition for this to be the case provides a s...
متن کاملScalable Dissipative Preparation of Many-Body Entanglement.
We present a technique for the dissipative preparation of highly entangled multiparticle states of atoms coupled to common oscillator modes. By combining local spontaneous emission with coherent couplings, we engineer many-body dissipation that drives the system from an arbitrary initial state into a Greenberger-Horne-Zeilinger state. We demonstrate that using our technique highly entangled ste...
متن کاملCellular Genetic Algorithm on Graphic Processing Units
The availability of low cost powerful parallel graphic cards has estimulated a trend to implement diverse algorithms on Graphic Processing Units (GPUs). In this paper we describe the design of a parallel Cellular Genetic Algorithm (cGA) on a GPU and then evaluate its performance. Beyond the existing works on masterslave for fitness evaluation, we here implement a cGA exploiting data and instruc...
متن کاملHermes: Scalable and Load Distribution Engine for General-Purpose Computing on Graphics Processing Units (GPGPU)
Modern Graphics Processing Units (GPU) constitute typical examples of multicore systems with more than one hundred processing. However, in order to efficiently exploit the increasing number of on-chip processing cores, a scalable enabler of massively parallel computing is imperative. This paper proposes Hermes, a scalable and dynamic load distribution engine that exploits hardware aggressively ...
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: Computer Physics Communications
سال: 2022
ISSN: ['1879-2944', '0010-4655']
DOI: https://doi.org/10.1016/j.cpc.2022.108472