2,775 Hits in 10.5 sec

Parallelization, Spatial Decomposition and Load Balancing of a Single Tree Level Forest Dynamics Simulator

Artur Signell, Johan Schöring, Mats Aspnäs, Jan Westerholm
2010 Mathematical and Computational Forestry & Natural-Resource Sciences  
SPATE-HPC is a single tree level forest dynamics simulator capable of simulating very large forest areas.  ...  We describe the domain decomposition methods and the load balancing strategy we have used to ensure good performance and scalability of huge simulations.  ...  This work received support from the Academy of Finland in the "Sustainable production and products" research programme KETJU.  ... 
dblp:journals/mcfns/SignellSAW10 fatcat:ev7lizyxlnb2xn2jctzj4vbnaa

Hybrid MPI-OpenMP Paradigm on SMP Clusters: MPEG-2 Encoder and N-Body Simulation [article]

Truong Vinh Truong Duy, Katsuhiro Yamazaki, Kosai Ikegami, Shigeru Oyanagi
2012 arXiv   pre-print
Similarly, the work load of the force calculation which accounts for upwards of 90% of the cycles in typical computations in the n-body simulation is shared among OpenMP threads after ORB domain decomposition  ...  SMP nodes and OpenMP for parallelism within pictures at the slice level within each SMP node.  ...  Exploiting Two Levels of Parallelism The hierarchical image data structure of MPEG bit-stream is eminently suitable for using the hybrid paradigm to apply parallelism beyond a single level.  ... 
arXiv:1211.2292v1 fatcat:wqiucpiacneqjeabuwbagvlqjq

4D-SAS: A Distributed Dynamic-Data Driven Simulation and Analysis System for Massive Spatial Agent-Based Modeling

Zhenqiang Li, Xuefeng Guan, Rui Li, Huayi Wu
2016 ISPRS International Journal of Geo-Information  
To evaluate the system's efficiency and scalability, two different spatially explicitly agent-based models, an en-route choice model, and a forest fire propagation model, were created on 4D-SAS.  ...  It can automatically decompose simulation tasks and distribute them among computing nodes following two common schemes: order division or spatial decomposition.  ...  Thus, it is important to integrate these two decomposition methods into a single simulation platform.  ... 
doi:10.3390/ijgi5040042 fatcat:ktcrkf2ysvdg3fyh3utgkf7qni

Parallel level-set methods on adaptive tree-based grids

Mohammad Mirzadeh, Arthur Guittet, Carsten Burstedde, Frederic Gibou
2016 Journal of Computational Physics  
Interestingly, tree-based grids have a nice spatial ordering that naturally leads to the concept of space-filling curves (SFCs) and can be efficiently exploited for parallel load balancing [4, 21, 14]  ...  More recently, the authors in [58] used a domain decomposition strategy on a cubed sphere but with a single layer of ghost nodes.  ...  This work used the Extreme Science and Engineering Discovery Environment (XSEDE), which is supported by National Science Foundation  ... 
doi:10.1016/ fatcat:n7jzm42gqbgubmubsgracy4jxa

Solving Problems on Parallel Computers by Cellular Programming [chapter]

Domenico Talia
2000 Lecture Notes in Computer Science  
We introduce the main constructs of CARPET and discuss how the language can be effectively utilized to implement natural solvers of real-world complex problems such as forest fire and circuitry simulations  ...  Performance figures of the experiments carried out on a MIMD parallel computer show the effectiveness of our approach both in terms of execution time and speedup.  ...  It can also be used to solve parallel computing problems such as routing strategies, task scheduling and load balancing, parallel computer graphics and cryptography.  ... 
doi:10.1007/3-540-45591-4_81 fatcat:uotdc3hpuvg25nqsqyicxtfsx4

Adaptive kinetic-fluid solvers for heterogeneous computing architectures

Sergey Zabelok, Robert Arslanbekov, Vladimir Kolobov
2015 Journal of Computational Physics  
Using GPUs enables hybrid simulations of mixed rarefied-continuum flows with a million of Boltzmann cells each having a 24x24x24 velocity mesh.  ...  Challenges are due to the irregular data access for adaptive Cartesian mesh, vast difference of computational cost between kinetic and fluid cells, and desire to evenly load all CPUs and GPUs during grid  ...  We wish to thank Dr Martin Burtscher for useful discussions and suggestion of the warp algorithm for the LBM solver. Thanks to an anonymous reviewer for useful suggestions for improving the paper.  ... 
doi:10.1016/ fatcat:cce7msllzbfhbp5jttoekob7su

Parallel simulation of multiphase flows using octree adaptivity and the volume-of-fluid method

Gilou Agbaglah, Sébastien Delaux, Daniel Fuster, Jérôme Hoepffner, Christophe Josserand, Stéphane Popinet, Pascal Ray, Ruben Scardovelli, Stéphane Zaleski
2011 Comptes rendus. Mecanique  
The impact of a droplet on a layer of the same liquid results in the formation of a thin air layer trapped between the droplet and the liquid layer that the adaptive refinement allows to capture.  ...  The parallelisation of Gerris is achieved by domain decomposition. We show examples of the capabilities of Gerris on several types of problems.  ...  Load balancing Combining adaptivity and parallelism makes the use of load-balancing techniques mandatory in order to maintain good parallel efficiency.  ... 
doi:10.1016/j.crme.2010.12.006 fatcat:yanhqyazhjgrjig32fci4d36du

Review of Nitrogen Fate Models Applicable to Forest Landscapes in the Southern U.S

2013 American Society of Agricultural and Biological Engineers. Transactions  
The field-scale models DRAINMOD-FOREST and REMM contain process-level components characterizing hydrology, forest growth, and N dynamics, but they have limited capability to describe transport processes  ...  Assessing the environmental impacts of fertilizer nitrogen (N) used to increase productivity in managed forests is complex due to a wide range of abiotic and biotic factors affecting its forms and movement  ...  ., and the three anonymous reviewers of this manuscript for their constructive suggestions and comments that enhanced the quality of the manuscript.  ... 
doi:10.13031/trans.56.10096 fatcat:xmxm3mgzqfdbrkrnadwz6qi4py

Cluster Optimization and Parallelization of Simulations with Dynamically Adaptive Grids [chapter]

Martin Schreiber, Tobias Weinzierl, Hans-Joachim Bungartz
2013 Lecture Notes in Computer Science  
We discuss runtime improvements resulting from the skip mechanism and the implications on shared memory performance and load balancing.  ...  While stationary clusters already can be handled more efficiently than their dynamic counterparts, we propose to treat them as atomic grid entities and introduce a skip mechanism that allows the grid traversal  ...  UK-c0020, made by the King Abdullah University of Science and Technology (KAUST). All software is freely available at  ... 
doi:10.1007/978-3-642-40047-6_50 fatcat:iq7lli2dlbeofdy72elibgh6lq

Electrostatic PIC with adaptive Cartesian mesh

Vladimir Kolobov, Robert Arslanbekov
2016 Journal of Physics, Conference Series  
It is shown that cell-based AMR provides a convenient particle management algorithm for exponential multiplications of electrons and ions in the ionization events.  ...  Challenges of PIC method with cell-based adaptive mesh refinement (AMR) are related to a decrease of the particle-per-cell number in the refined cells with a corresponding increase of the numerical noise  ...  Acknowledgments This work was partially supported by the DARPA SBIR Project W31P4Q-15-C-0047 and by the US Department of Energy Office of Fusion Energy Science Contract DE-SC0001939.  ... 
doi:10.1088/1742-6596/719/1/012020 fatcat:nnj2suisu5dwhob3slhrvei6lm

Adaptive Kinetic-Fluid Solvers for Heterogeneous Computing Architectures [article]

Sergey Zabelok, Robert Arslanbekov, Vladimir Kolobov
2015 arXiv   pre-print
Using GPUs enables hybrid simulations of mixed rarefied-continuum flows with a million of Boltzmann cells with 24x24x24 velocity mesh.  ...  We describe the implementation of CUDA kernels for three modules in UFS: the direct Boltzmann solver using discrete velocity method, the Direct Simulation Monte Carlo (DSMC) solver, and a mesoscopic solver  ...  We wish to thank Dr Martin Burtscher for useful discussions and suggestion of the warp algorithm for the LBM solver.  ... 
arXiv:1503.00707v2 fatcat:4io4w5otxzd3dd4kljukfn5xim

Extreme-Scale AMR

Carsten Burstedde, Omar Ghattas, Michael Gurnis, Tobin Isaac, Georg Stadler, Tim Warburton, Lucas Wilcox
2010 2010 ACM/IEEE International Conference for High Performance Computing, Networking, Storage and Analysis  
Due to the complex dynamic data structures and frequent load balancing, scaling dynamic AMR to hundreds of thousands of cores has long been considered a challenge.  ...  Many problems are characterized by dynamics occurring on a wide range of length and time scales.  ...  We thank Laura Alisic, George Biros, Martin Burtscher, Rahul Sampath, and Tiankai Tu for extended discussions.  ... 
doi:10.1109/sc.2010.25 dblp:conf/sc/BursteddeGGISWW10 fatcat:j2igjrbui5hf3a4rgeki73ltva

Modeling vegetation structure-ecosystem process interactions across sites and ecosystems

W.K. Lauenroth, D.L. Urban, D.P. Coffin, W.J. Parton, H.H. Shugart, T.B. Kirchner, T.M. Smith
1993 Ecological Modelling  
traits of grasses and trees.  ...  Preliminary results indicate that life-history traits of plants have a profound influence on system-level behaviors, and that differences between grasslands and forests can be attributed largely to contrasting  ...  We appreciate the enthusiasm and assistance of collaborators at the several LTER sites.  ... 
doi:10.1016/0304-3800(93)90099-e fatcat:ut3pod2cenbdtfu5y3xvwdlwpa

Parallel TreeSPH

Romeel Davé, John Dubinski, Lars Hernquist
1997 New Astronomy  
A cosmological simulation from z=49 to z=2 with 64^3 gas particles and 64^3 dark matter particles requires ~6000 node-hours on a Cray T3D, with a communications overhead of ~10% and is load balanced to  ...  PTreeSPH utilizes a domain decomposition procedure and a synchronous hypercube communication paradigm to build self-contained subvolumes of the simulation on each processor at every timestep.  ...  We acknowledge grants of computer resources by Pittsburgh Supercomputing Center and the Cornell Theory Center.  ... 
doi:10.1016/s1384-1076(97)00019-5 fatcat:veqxsydvy5fvxgrt223m4kx4gm

Load Balancing and Data Locality in Adaptive Hierarchical N-Body Methods: Barnes-Hut, Fast Multipole, and Radiosity

J.P. Singh, C. Holt, T. Totsuka, A. Gupta, J. Hennessy
1995 Journal of Parallel and Distributed Computing  
We find that straightforward decomposition techniques which an automatic scheduler might implement do not scale well, because they are unable to simultaneously provide load balancing and data locality.  ...  In this paper, we study the partitioning and scheduling techniques required to obtain effective parallel performance on applications that use a range of hierarchical N-body applications.  ...  Dennis Roger and Eric Bruni developed an initial implementation of the parallel code as a class project at Stanford University.  ... 
doi:10.1006/jpdc.1995.1077 fatcat:xc6rf5l73rchndyvjx6sct7k3i
« Previous Showing results 1 — 15 out of 2,775 results