This project is to build a vector processing version of pcgpak, the preconditioned conjugate gradient package of subroutines for iterative solution of large, sparse, nonsymmetric systems of linear equations. the pcgpak is currently being marketed and is the result of internal development work… More

We propose to build a state-of-the-art software package for solving general large systems of nonlinear equations. although there are several software packages for solving large sparse linear systems (sparsepack, ysmp, harewell, for example), similar packages for nonlinear systems are almost… More

Modifications of existing software employed in the numerical modeling of multidimensional reacting fuel flows are proposed which would improve the computational efficiency and robustness of such methods and allow the inclusion of a more detailed representation of the chemical and physical processes… More

The software available for use on supercomputers has provided neither the most sophisticated mathematical techniques available nor a particularly convenient working environment for the application programmer. as a result, the extremely scarce and expensive supercomputing capabilitythat is available… More

Message-passing architectures (such as the fpst series, the ncube 10, the intel ipsc, and thinking machines connection machine) are one of the two major approaches to building parallel architectures. these architectures have the potential to incorporate large numbers (greater than 100) of… More

Shared-memory architectures (e.g. the cray-2, the cray-xmp, the encore multimax, the sequent balance series, the bbn butterfly, and the ibm rp3 project) are one of the two major approaches to building parallel architectures. for a limited number of processors, the interprocessor communication cost… More

Hypercube-connected multicomputers like the intel ipsc are of growing interest to the parallel programming community. they offer advantages in scalability and in cost over competing multi-computer architectures. we propose to buildan implementation of the parallel programming language lindaon the… More

Crystal and the crystal run-time system will make the use of large-scale multiprocessors practical for real-world problems. crystal is a very-high-level language in which a user program resembles a concise, formal mathematical description. crystal is easy to use yet has the modularity and freedom… More

For the rest of this century, two of the potentially most important advances in the "technology" of large scale scientific computing are sparse matrix (iterative) algorithms and massively parallel processing. in this proposal, we outline our plans to build a software package for the connection… More

Our outlined plans to build a vector processing version of pcgpaktm, the preconditioned conjugate gradient package of subroutines for interative solution of large, sparse, nonsymmetric systems of linear equations. pcgpak, which is currently being marketed by scientific computing associates,… More

Significant numerical computations now commonly require vastprocessing power. at the core of many of these problems is the solution of large, sparse systems of equations. a host of specialized high-speed processors have been developed to attack large problems. systolic architectures have proven to… More

Iterative methods for solving sparse linear systems require significantly fewer arithmetic operations and less memory than currently popular direct methods. however, for very large problems, for example as derived from discretizations of partial differential equations in three space variables, it is… More

We propose to develop a portable, parallel library of high-level basic linear algebraic subprograms for use by the national scientific community. such a portable parllel library of routines will aid greatly in the dissemination and widespread use of advanced linear algebra algorithms on a variety of… More

The main impediment to the use of parallel and distributed computing is the lack of proper high-level parallel programming environment. the potential of parallel machines will never be tapped if they cnanot be programmed effectively and easily. fortran-linda is a programming language for building… More

In the last few years, the performance of microprocessors used in workstations has increased dramaticlly, and advancesin vlsi technology assure continuing performance improvement in the future. scientific computing associates, inc. has made significant progress in the development of an efficient… More

We propose to integrate a powerful interactive software environment for scientific computing and visualization with suites of high-quality numerical subroutine libraries, including parallel and vector routines. the advantages of the proposed system include rapid and efficient prototyping of… More

Global optimization plays a key role in molecular computations in computational chemistry. due to the large size of even modest molecules, these methods must contend with countless local minima. clever algorithms have been developed that sample parameter space to reduce trapping in local minima, but… More

Management and search of large databases is a computational bottleneck in many domains. one attractive approach to eliminating it is to exploit recent advances in parallel computer hardware and processing. however, the development of software that permits this represents one of the major challenges… More

Real-time monitors are designed to filter and analyze patterns in collections of incoming data streams--for example, to convert low-level signal and image data into high-level descriptions. in short, they are a solution to the "data fusion" problem. this project implements the process-trellis… More

Large multi-user databases are already one of the most significant applicatons of computers, and their importance will only grow as we enhance our abilities to automatically generate and capture huge amounts of data. users increasingly will demand access to very large distributed knowledge bases… More

We propose to develop an extensible, cost effective, parallel supercanputer based on low cost, high performance muitlprocessors connected via a hlgh-performance local area network. most of our effort will focus on multiprocessors which are merchant, shared memory systems, though the general ideas… More

Math libraries are an important part of programming environments that support the development of application software. while broadly available on sequential computers, math libraries are rare for parallel computers-especially in the case of distributed computing across networks of workstations.… More

A "hypercomputer" is the parallel computer that emerges when one sums the unused cycles over all the computer nodes on a local area network. however, fault-tolerance is a critical issue for longer-lived applications running on typical lans where node failures are not uncommon. process checkpoint and… More

This research explores the issues involved in parallelizing gaussian 92, the most widely used electronic structure software package, using the linda programming environment for workstation networks and multiprocessors. since both gaussian 92 and linda emphasize portability across multiple computer… More

As a National information resource, the Internet makes available vast amounts of information to a broad range of consumers. For most users, however, only a tiny fraction of the aggregate information is significant, and the overall resource only has value if that fraction can be located reliably,… More

Object Oriented Programming (OOP) techniques are a vital part of modern software engineering. Unfortunately, on parallel computers where the need is greatest, OOP tools are lacking. We propose to address this problem by designing a parallel OOP language. Unlike other parallel OOP research efforts,… More

Many aerospace applications produce vast amounts of data that must be analyzed intelligently as rapidly as possible, often with near- real-time constraints. Examples include telemetry data from a variety of sources, scientific data from space missions, and design and performance data for complex… More

This project addresses the important area of programming support for high performance parallel computing. At present, message-passing systems like PVM or MPI are the most widely used software systems for parallel computing, both for defense applications and for applications in the commercial and… More

Extracting knowledge from the vast amount of rapidly-evolving data handled by image processing programs, situation assessment and decision support systems, and many other applications important to the BMDO mission is a demanding task requiring powerful computer hardware and appropriate software… More

Parallel and distributed processing on "metacomputers" formed by networking scalable parallel computers with COTS workstations and PCs is clearly the wave of the future for high-performance computing. The reasons are obvious: cost and performance. The Department of Defence can benefit greatly,… More

This SBIR Phase I project will investigate the use of virtual shared memory technology as a unifying software paradigm for a cluster of shared-memory multiprocessors interconnected on a local area network or on a high-speed communication fabric. The main focus of the research will be on development… More

This SBIR Phase I project will investigate the use of virtual shared memory technology as a unifying software paradigm for a cluster of shared-memory multiprocessors interconnected on a local area network or on a high-speed communication fabric. The main focus of the research will be on development… More

51127-98-I A Java-Based Environment for Adaptive, Fault-Tolerant Parallel and Distributed Computing--Scientific Computing Associates, Inc., One Century Tower, 265 Church Street, New Haven, CT 06510-7013; (203) 777-7442 Dr. Robert Bjornson, Principal Investigator Dr. Andrew H. Sherman, Business… More

Modern methods of drug discovery rely heavily on computer modeling of potential therapeutic agents to calculate their geometric and electronic properties and optimize them to interact specifically with target receptors with high potency. This process is extremely compute intensive due to the large… More