The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
Efficient and elegant complex numbers are one of the preconditions for the use of Java in scientific computing. This paper introduces a preprocessor and its translation rules that map a new basic type complex and its operations to pure Java. For the mapping is insufficient to just replace one complex-variable with two double-variables. Compared to code that uses Complex objects and method invocations...
The executable data structures method is a kind of run-time code generation. It reduces the traversal time of data structures by implementing the traversal operation as self-traversal code stored in each data structure. It is attractive but not easy to use. This paper shows that it can be viewed as run-time optimization of a group of objects and can automatically be generated from the source program...
Application codes reliably achieve performance far less than the advertised capabilities of existing architectures, and this problem is worsening with increasingly-parallel machines. For large-scale numerical applications, stencil operations often impose the greater part of the computational cost, and the primary sources of inefficiency are the costs of message passing and poor cache utilization....
Communication networks have steadily increased in size and complexity to meet the growing demands of applications. Simulations have been used to model and analyze modern communication networks. Modeling and simulation of networks involving thousands of nodes is hard due to their sheer size and complexity. Complete models of the ultra-large networks need to be simulated in order to conduct in-depth...
We introduce a scalable, extensible object-oriented system developed primarily for signal processing and synthesis for musical and multimedia applications. The main performance issue with these applications concerns functions of discrete-time. Novel techniques exploit fine-grain parallelism in the calculation of these functions to allow users to express them at a high-level in C++. New scheduling...
In hospital services, practitioners need to access and study large data volume (3D images) with help of specific, parallel, high performance processing tools. This work describes the ARAMIS platform (A Remote Access Medical Imaging System), which allows transparent remote access to parallel image processing libraries. Such system is based on a communication protocol which takes as input parallel libraries...
Component technologies offer a promising approach for managing the increasing complexity and interdisciplinary nature of high-performance scientific applications. Language interoperability is required for components written in different languages to communicate. In this paper, we present an approach to language interoperability for high-performance parallel components. Based on Interface Definition...
Visper is a network-based, visual software-engineering environment for parallel processing. It is completely implemented in Java and supports the message-passing model. Java offers the basic platform independent services needed to integrate heterogeneous hardware into a seamless computational resource. Easy installation, participation and flexibility are seen as the key properties when using the system...
None of the current attempts to provide an Internet-wide global computing infrastructure presents well-defined programming constructs such as object distribution, dispatching, migration and concurrency with maximum portability and high transparency to a programmer. We propose a Web-based global computing infrastructure called Tiger, providing well-defined object-oriented programming constructs They...
SIFFEA is an automated system for parallel finite element method (PFEM) with unstructured meshes on distributed memory machines. It synthesizes mesh generator, mesh partitioner, linear system assembler and solver, and adaptive mesh refiner. SIFFEA is an implicit parallel environment: The user need only specify the application model in serial semantics; all internal communications are transparent to...
The Overture framework is an object-oriented environment for solving partial differential equations in two and three space dimensions. It is a collection of C++ libraries that enables the use of finite difference and finite volume methods at a level that hides the details of the associated data structures. Overture can be used to solve problems in complicated, moving geometries using the method of...
We use the paradigm of generic programming together with the Standard Template Library (STL) to develop concepts and template classes that can represent finite sets and their relations as they occur in (parallel) finite element and other numerical methods. A key idea of our approach is to consider these sets as static search structures with clearly separated phases for insertions and retrievals. The...
Fill-reducing sparse matrix orderings have been a topic of active research for many years. Although most such algorithms are developed and analyzed within a graph-theoretical framework, for reasons of performance the corresponding implementations are typically realized with programming languages devoid of language features necessary to explicitly represent graph abstractions. Recently, generic programming...
The concept of design patterns has recently emerged as a new paradigm in the context of object-oriented design methodology. Similar ideas are being explored in other areas of computing. In the parallel computing domain, design patterns describe recurring parallel computing problems and their solution strategies. Starting with the late 1980’s, several pattern-based systems have been built for facilitating...
Modern computational fluid dynamics (CFD) software is complex. Often CFD simulations require complex geometries, flexible boundary conditions, multiple integrated computational models (for example, heat conduction, structural deformations, gas dynamics, etc.), as well as grid adaptation. As a result of this complexity, the correct implementation of numerical simulation components is actually less...
OptSolve++ is a set of C++ class libraries for nonlinear optimization and root-finding. The primary components include TxOptSlv (optimizer and solver classes), TxFunc (functor classes used to wrap user-defined functions), TxLin (linear algebra), and a library of test functions. These cross-platform components use exception handling and encapsulate diagnostic output for the calling application. Use...
The investigation of strong quantum effects is one of the major research areas in physics with important application aspects, like the high temperature superconductors. The only reliable approaches to these systems are large scale numerical simulations, made difficult by the exponential scaling of most algorithms. We present efforts in developing a generic and parallel program package for the simulation...
Applications executing on complex computational systems provide a challenge for the development of runtime performance monitoring software. We discuss a computational model, application monitoring, data access models, and profiler functionality. We define data consistency within and across threads as well as across contexts and nodes. We describe the TAU runtime monitoring framework which enables...
This paper presents how a specific technique (SharedOnRead) can be used to improve performances of a distributed objects language. We present results for a network of workstations, providing comparisons between MPI and C++// implementations, both on homogeneous and heterogeneous platforms. The results could be applied to other models of distribution, such as RMI.
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.