Performance Modeling of Unstructured Mesh Particle Transport Computations PDF Download

Author:
Publisher:
ISBN:
Category :
Languages : en
Pages : 8

View

Book Description
The performance of unstructured mesh applications presents a number of complexities and subtleties that do not arise for dense structured meshes. From a prograniming point of view, handling an unstructured mesh has an increased complexity to manage the necessary data structures and interactions between mesh-cells. From a performance point of view, there are added dificulties in understanding both the processing time on a single processor and the scaling characteristics. In this work we present a performance model for the calculation of deterministic SN transport on unstructured meshes. It builds upon earlier work that successfully modeled the same calculation on structured meshes. The model captures the key processing characteristics and is parametric using both the system performance data (latency, bandwidth, processing rate etc.) and application data (mesh size etc.) as input. The model is validated on two clusters [an HP Alphaserver and an Itanium-2 system] showing high accuracy. Importantly it is also shown that a single formulation of the model can be used to predict the performance of two quite different implementations of the same calculation.

Author: Peter M.A. Sloot
Publisher: Springer
ISBN: 3540448632
Category : Computers
Languages : en
Pages : 1183

View

Book Description
Some of the most challenging problems in science and engineering are being addressed by the integration of computation and science, a research ?eld known as computational science. Computational science plays a vital role in fundamental advances in biology, physics, chemistry, astronomy, and a host of other disciplines. This is through the coordination of computation, data management, access to instrumentation, knowledge synthesis, and the use of new devices. It has an impact on researchers and practitioners in the sciences and beyond. The sheer size of many challenges in computational science dictates the use of supercomputing, parallel and distri- ted processing, grid-based processing, advanced visualization and sophisticated algorithms. At the dawn of the 21st century the series of International Conferences on Computational Science (ICCS) was initiated with a ?rst meeting in May 2001 in San Francisco. The success of that meeting motivated the organization of the - cond meeting held in Amsterdam April 21–24, 2002, where over 500 participants pushed the research ?eld further. The International Conference on Computational Science 2003 (ICCS 2003) is the follow-up to these earlier conferences. ICCS 2003 is unique, in that it was a single event held at two di?erent sites almost opposite each other on the globe – Melbourne, Australia and St. Petersburg, Russian Federation. The conference ran on the same dates at both locations and all the presented work was published in a single set of proceedings, which you hold in your hands right now.

Author: Sang-Soo Yeo
Publisher: Springer Science & Business Media
ISBN: 3642131182
Category : Computers
Languages : en
Pages : 596

View

Book Description
This book constitutes the proceedings of the 10th International Conference on Algorithms and Architectures for Parallel Processing, ICA3PP. The 47 papers were carefully selected from 157 submissions and focus on topics for researchers and industry practioners to exchange information regarding advancements in the state of art and practice of IT-driven services and applications, as well as to identify emerging research topics and define the future directions of parallel processing.

Author: Vladimir Getov
Publisher: Springer Science & Business Media
ISBN: 1461503612
Category : Computers
Languages : en
Pages : 290

View

Book Description
Past and current research in computer performance analysis has focused primarily on dedicated parallel machines. However, future applications in the area of high-performance computing will not only use individual parallel systems but a large set of networked resources. This scenario of computational and data Grids is attracting a great deal of attention from both computer and computational scientists. In addition to the inherent complexity of parallel machines, the sharing and transparency of the available resources introduces new challenges on performance analysis, techniques, and systems. In order to meet those challenges, a multi-disciplinary approach to the multi-faceted problems of performance is required. New degrees of freedom will come into play with a direct impact on the performance of Grid computing, including wide-area network performance, quality-of-service (QoS), heterogeneity, and middleware systems, to mention only a few.

Author: John Levesque
Publisher: CRC Press
ISBN: 1420077066
Category : Computers
Languages : en
Pages : 244

View

Book Description
High Performance Computing: Programming and Applications presents techniques that address new performance issues in the programming of high performance computing (HPC) applications. Omitting tedious details, the book discusses hardware architecture concepts and programming techniques that are the most pertinent to application developers for achievi

Author:
Publisher:
ISBN:
Category : Computers
Languages : en
Pages : 622

View

Book Description
Workshop on Web Services-based Grid Applications (WSGA) -- Workshop on Parallel and Distributed Multimedia (PDM), the Workshop on Wireless and Sensor Networks (WSNet) -- 3rd International Workshop on Embedded Computing (EC-06) -- Workshop on Performance Evaluation of Networks for Parallel, Cluster and Grid Computing Systems (PEN-PCGCS) -- the 5th Workshop on Compile and Runtime Techniques for Parallel Computing, (CRTPC) -- 8th Workshop on High Performance Scientific and Engineering Computing (HPSEC-06)

Author:
Publisher:
ISBN:
Category :
Languages : en
Pages : 10

View

Book Description
This paper addresses two key parallelization challenges the unstructured mesh-based ocean modeling code, MPAS-Ocean, which uses a mesh based on Voronoi tessellations: (1) load imbalance across processes, and (2) unstructured data access patterns, that inhibit intra- and inter-node performance. Our work analyzes the load imbalance due to naive partitioning of the mesh, and develops methods to generate mesh partitioning with better load balance and reduced communication. Furthermore, we present methods that minimize both inter- and intranode data movement and maximize data reuse. Our techniques include predictive ordering of data elements for higher cache efficiency, as well as communication reduction approaches. We present detailed performance data when running on thousands of cores using the Cray XC30 supercomputer and show that our optimization strategies can exceed the original performance by over 2×. Additionally, many of these solutions can be broadly applied to a wide variety of unstructured grid-based computations.

Author: Liangzhi Cao
Publisher: Woodhead Publishing
ISBN: 0128182210
Category : Technology & Engineering
Languages : en
Pages : 292

View

Book Description
Deterministic Numerical Methods for Unstructured-Mesh Neutron Transport Calculation presents the latest deterministic numerical methods for neutron transport equations (NTEs) with complex geometry, which are of great demand in recent years due to the rapid development of advanced nuclear reactor concepts and high-performance computational technologies. This book covers the wellknown methods proposed and used in recent years, not only theoretical modeling but also numerical results. This book provides readers with a very thorough understanding of unstructured neutron transport calculations and enables them to develop their own computational codes. The fundamentals, numerical discretization methods, algorithms, and numerical results are discussed. Researchers and engineers from utilities and research institutes are provided with examples on how to model an advanced nuclear reactor, which they can then apply to their own research projects and lab settings. Combines the theoretical models with numerical methods and results in one complete resource Presents the latest progress on the topic in an easy-to-navigate format

Author:
Publisher:
ISBN:
Category :
Languages : en
Pages : 20

View

Book Description
The majority of finite element models in structural engineering are composed of unstructured meshes. These unstructured meshes are often very large and require significant computational resources; hence they are excellent candidates for massively parallel computation. Parallel solution of the sparse matrices that arise from such meshes has been studied heavily, and many good algorithms have been developed. Unfortunately, many of the other aspects of parallel unstructured mesh computation have gone largely ignored. The authors present a set of algorithms that allow the entire unstructured mesh computation process to execute in parallel -- including adaptive mesh refinement, equation reordering, mesh partitioning, and sparse linear system solution. They briefly describe these algorithms and state results regarding their running-time and performance. They then give results from the 512-processor Intel DELTA for a large-scale structural analysis problem. These results demonstrate that the new algorithms are scalable and efficient. The algorithms are able to achieve up to 2.2 gigaflops for this unstructured mesh problem.

Author:
Publisher:
ISBN:
Category : Aeronautics
Languages : en
Pages : 652

View

Book Description
Lists citations with abstracts for aerospace related reports obtained from world wide sources and announces documents that have recently been entered into the NASA Scientific and Technical Information Database.