Scheduled System Maintenance
On Saturday, October 21, single article sales and account management will be unavailable until 6 PM ET.
Notice: There is currently an issue with the citation download feature. Learn more.

CompEuro 1992 Proceedings Computer Systems and Software Engineering

4-8 May 1992

Filter Results

Displaying Results 1 - 25 of 126
  • Parallel routing algorithms for incomplete hypercube interconnection networks

    Publication Year: 1992, Page(s):418 - 423
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (483 KB)

    A severe restriction on a complete hypercube topology is that the number of processors must be a power of 2. Such a restriction limits its applicability. An incomplete hypercube may provide more flexibility in system sizes. There are three classes of variant incomplete hypercube interconnection networks: I/sup r//sub m/, I/sup r//sub M/ and I/sup r//sub A/. The authors focus on the parallel routin... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Genesis: a generic simulation subsystem for parallel architectures

    Publication Year: 1992, Page(s):412 - 417
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (435 KB)

    Genesis is a library of C++ classes that serves as building blocks in modeling parallel architectures. This paper presents the architectural representation scheme on which Genesis is based. The design choices of Genesis and some of the reasons behind the choices are discussed. The author describes the way the performance parameters (architecture, program, assignment method, and routing scheme) are... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A generalized stochastic Petri net model of Multibus II

    Publication Year: 1992, Page(s):406 - 411
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (440 KB)

    Results on the performance of a Multibus II based automation system are presented. These are achieved by means of a generalized stochastic Petri net model which was solved by analytical techniques instead of by simulation. The mapping of this industry-standard protocol to a stochastic Petri net model is outlined. It is shown how protocol features like arbitration priorities, parking, and message p... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Distributed system modeling with bidirectional Petri nets

    Publication Year: 1992, Page(s):401 - 405
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (343 KB)

    A new tool for modeling systems called bidirectional Petri nets (BPNs) is presented. It increases the degree of abstraction drastically, and makes it possible to model complicated systems concisely. The method uses a two-level system modeling technique. The first level has the responsibility for all basic transmissions, and the second level models the higher-level structure of the whole system. Th... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Validation using CASE tools

    Publication Year: 1992, Page(s):389 - 394
    Cited by:  Patents (3)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (271 KB)

    Describes how computer assisted software engineering (CASE) tools were used in the process of validation of several reactor physics codes. ARCHY is a tool that automates most of the drudgery involved in making a programmer's reference manual. When FORTRAN source code is read by ARCHY, it outputs a data dictionary that includes a categorization of variables and common block information. It also pri... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Constructing a multitasking programming environment

    Publication Year: 1992, Page(s):383 - 388
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (434 KB)

    The construction of a multitasking programming environment called MCPE is presented. MCPE contains multiple tasks so that it can speed up response and utilize idle CPU time. Tasks communicate with each other through shared data storage. The tasks (such as a structure editor and a static analyzer) were designed as lightweight processes. The shared data storage (such as the grammar tree and the symb... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • An architecture for active hypertext on distributed systems

    Publication Year: 1992, Page(s):377 - 382
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (495 KB)

    The combination of hypertext and distributed systems technologies enhances effectiveness in information management and dissemination. An architecture which integrates these technologies is presented, the issues that influenced its design are cited, and its realization is outlined. Easy and graceful integration of new applications in the framework is supported by the generic and extensible characte... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • VAMP: A tool for literate programming independent of programming language and formatter

    Publication Year: 1992, Page(s):371 - 376
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (548 KB)

    The authors have developed a module extractor called VAMP, that cooperates with a standard formatter. They describe the VAMP approach to literate programming. Literate programming identifies a method of documentation which explains to people what a computer is supposed to do. Generally, this means that refinement steps are documented in such a way that modules can be extracted from the documentati... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Constructing multi-user applications in REX

    Publication Year: 1992, Page(s):365 - 370
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (407 KB)

    Presents a configuration language based approach to the description and construction of distributed multiuser applications. The authors describe the facilities of the configuration language DARWIN which permit the description of multiuser applications in a manner identical to normal configuration-based parallel or distributed programs. The implementation platform, REX, is described, as is Conf, a ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Memory bank conflict on some vector supercomputers

    Publication Year: 1992, Page(s):359 - 364
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (388 KB)

    The effects of memory bank conflict on some typical vector supercomputers were investigated. The numerical experiments were carried out on the NEC SX-2, the Hitachi S-820/80, and the Fujitsu VP-2600/20 and in part on an NEC SX-3. The deterioration of performance due to memory bank conflict is studied, and a recipe is given for choosing the numbers of gridpoints in the x, y and z directions for eff... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Hiding communication latencies in message-passing multiprocessors: a case study based on neural network simulation

    Publication Year: 1992, Page(s):353 - 358
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (445 KB)

    An application package which allows the user to explore the possibility of hiding communication latencies in message-passing multiprocessors is examined. The package is a neural network simulator for transputer-based machines. Simply by modifying few high-level directives in a configuration file, heuristic strategies for fully overlapping communication delays with internal computation, can be easi... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • HiTest: an architecture for highly parallel software

    Publication Year: 1992, Page(s):347 - 352
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (482 KB)

    HiTest is a special-purpose, dynamically reconfigurable, multiple single-instruction, multiple-data (SIMD) architecture designed for highly parallel software testing. The architecture can substantially reduce the total time needed for testing. It is particularly well suited for mutation analysis. The architecture and its use for mutation analysis and software testing in general are described, and ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • An associative memory approach to parallel logic event-driven simulation

    Publication Year: 1992, Page(s):341 - 346
    Cited by:  Patents (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (442 KB)

    Presents a parallel processing approach to logic simulation, called APPLES, in which gate evaluations and signal updating are executed in parallel in associative memory, rather than in the processor. This approach does not require any event scheduling mechanism and can model various logic gate types and delay models. Two concepts are coupled together to form a parallel acceleration technique in th... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Solution of matrix equations via algebraic transformations

    Publication Year: 1992, Page(s):336 - 340
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (199 KB)

    An algebraic result useful in the computational solution of sets of algebraic matrix equations is derived. An important advantage of the method is that the solution of only one equation is sufficient to determine the solutions of all matrix equations of the class considered without these equations having to be solved directly. An application example from the area of system theory is presented.< View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Parallel-form adaptive state-space filtering and its implementation

    Publication Year: 1992, Page(s):330 - 335
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (411 KB)

    Introduces an adaptive recursive state-space algorithm and its implementation. The adaptive state-space filtering is realized by using the least mean square adaptation algorithm and the gradients for filter parameters are derived directly from the state equations. To reduce the computational complexity, a parallel form of second-order sections is used. The performance of adaptive state-space filte... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Parallel architecture for a pel-recursive motion estimation algorithm

    Publication Year: 1992, Page(s):322 - 329
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (515 KB)

    Presents a parallel architecture for a pixel-recursive motion estimation algorithm. It is a linear array of processors, each consisting of an initialization part, a data-routing part, and an updating part. The initialization part performs a prediction of the motion vector. The routing part constitutes routing path along which previous frame data are routed from processors that store such data to p... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • The quotient SVD and directional aspects of 2 transfer matrices

    Publication Year: 1992, Page(s):316 - 321
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (368 KB)

    Directional aspects of multiple-input-multiple-output systems relative to one another are studied. The quotient singular value decomposition (QSVD), a generalization of the SVD is used. The definition and some properties of the QSVD are stated. The concepts of the indicator ellipsoid and the oriented energy plot allow for a better visualization of the directional gain. The oriented signal-to-signa... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • High-performance MCMs for mainframe computers

    Publication Year: 1992, Page(s):310 - 315
    Cited by:  Patents (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (628 KB)

    Presents a basic scheme for water-cooled multichip modules (MCMs). Multichip modules are a very effective approach to achieve high packaging density, excellent cooling and high electrical performance of mainframe processors. Two characteristic approaches are compared, namely an MCM with flip-chip mounting and an MCM with micropacks. The electrical characteristics of chip-to-board connections, micr... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Trends in cooling technology for high-density packaging

    Publication Year: 1992, Page(s):304 - 309
    Cited by:  Papers (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (353 KB)

    A general look at the development of cooling systems for multichip modules over the last few years reveals a clear increase in the levels of power dissipation. Heat transfer and cooling methods for multichip modules are considered. The extremely high demands placed on cooling systems are leading to increasingly cost-intensive technical solutions. Microcooling systems must be capable of integration... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Simulation of IC- and MCM-interconnects

    Publication Year: 1992, Page(s):298 - 303
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (412 KB)

    Different methods for the mathematical description of lines and line systems are analyzed and discussed. A simple method for the simulation of line systems, including losses, in a nonlinear circuit environment in the time domain is treated, and as an example a typical simulation tool (called ISIM) based upon this method is presented. The most crucial problem within the simulation of interconnects ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Micro electronic packaging technologies: status and trends

    Publication Year: 1992, Page(s):292 - 297
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (427 KB)

    Reviews the status of today's prevalent technologies for packaging microelectronic components for signal processing systems with an emphasis on high-performance computer applications. Some limits of standard hierarchical packaging schemes are discussed. Alternative novel chip packaging approaches like thin-film multichip carriers and direct chip attach methods are presented. Perspectives for some ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Simulation methods and technologies for multichip modules

    Publication Year: 1992, Page(s):286 - 291
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (333 KB)

    Concepts of multi-chip modules (MCMs) are discussed and simulation methods for their design are presented. The basic requirements for MCM materials and interconnection technologies are outlined. Signal transmission behavior and thermal problems are considered. A short survey of technologies for multilayers is given.<> View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Transient fault tolerance in redundant systems

    Publication Year: 1992, Page(s):280 - 285
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (499 KB)

    Outlines some problems which result from transient faults in designing fault tolerant systems. Based on this analysis, a systematic approach to effective transient fault tolerance in redundant systems is presented. The main point of this approach is to design error handling procedures, taking into account a variety of available error detection mechanisms and the context of error appearance. Implem... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Test quality improvement by physical testability enhancement

    Publication Year: 1992, Page(s):274 - 279
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (408 KB)

    A methodology that provides a way to control the test quality of VLSI systems by predicting, diagnosing, and improving the IC defect coverage is presented for the case of the physical implementation of boundary scan circuitry, together with the software tools that implement it. The method allows the identification of hard-to-detect faults, their physical origin and layout location, leading to sugg... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Design methodology for systematic derivation of fault-tolerant array processors

    Publication Year: 1992, Page(s):562 - 567
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (397 KB)

    A systematic approach for mapping of iterative algorithms into fault-tolerant processor arrays is presented. The initial description of the algorithms is Fortran-like nested loops, and the restrictions of the intermediate forms such as UREs are avoided. The principles of the coordinate method are used and regular or piecewise regular arrays can be derived. The allocation and the scheduling of the ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.