By Topic

Supercomputing '88:Proceedings of the 1988 ACM/IEEE Conference on Supercomputing, Vol. I

14-18 Nov. 1988

Filter Results

Displaying Results 1 - 25 of 58
  • Proceedings. Supercomputing '88 (IEEE Cat. No.88CH2617-9)

    Publication Year: 1988
    Request permission for commercial reuse | PDF file iconPDF (204 KB)
    Freely Available from IEEE
  • Faust: an environment for programming parallel scientific applications

    Publication Year: 1988, Page(s):3 - 10
    Cited by:  Papers (3)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (501 KB)

    The goals and architecture of the Faust environment are presented. The components of the Faust user-interface library are described. The Faust program abstraction supports three levels of detail: the process graph, the subroutine interconnection graph, and the program source code. Each level is described in detail.<> View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • An efficient pipelined dataflow processor architecture

    Publication Year: 1988, Page(s):368 - 373
    Cited by:  Papers (20)  |  Patents (7)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (428 KB)

    It is demonstrated that the principles of pipelined instruction execution can be effectively applied in data-flow computers, yielding an architecture that avoids the main sources of pipeline gaps during program execution in many conventional designs. The processing element uses an architecture called argument-fetch data-flow architecture. It has two parts: a data-flow instruction scheduling unit (... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • CRegs: a new kind of memory for referencing arrays and pointers

    Publication Year: 1988, Page(s):360 - 367
    Cited by:  Papers (7)  |  Patents (11)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (580 KB)

    Pointer and subscripted array references often touch memory locations for which there are several possible aliases; hence these references cannot be made from registers. Although conventional caches can increase performance somewhat, they do not provide many of the benefits of registers, and do not permit the compiler to perform many optimizations associated with register references. The CReg (pro... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • V-Pascal: an automatic vectorizing compiler for Pascal with no language extensions

    Publication Year: 1988, Page(s):182 - 189
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (660 KB)

    A detailed anatomy of the automatic vectorizing compiler V-Pascal (Version 1, now operational) is given. With no language extensions, V-Pascal efficiently vectorizes the whole of arbitrarily given multiply nested for loops using the mechanism of vector indirect addressing. Higher functions of V-Pascal, such as vectorization of WHILE-DO and recursive calls, vectorization of relational operations on... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • The design of a lockup-free cache for high-performance multiprocessors

    Publication Year: 1988, Page(s):352 - 359
    Cited by:  Papers (5)  |  Patents (15)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (680 KB)

    The performance of cache-based, shared-memory multiprocessors can suffer greatly from moderate cache miss rates because of the usually high ratio between memory-access and cache-access times. The authors propose a lockup-free cache design in which the handling of one or several cache misses is overlapped with processor activity. In multiprocessors, lockup-free caches aggravate the memory coherence... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Growing discord: programming philosophy and hardware design

    Publication Year: 1988, Page(s):18 - 26
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (688 KB)

    Current compiler technology is examined with respect to choices or trends in supercomputer design. The basic techniques are extensions of strategies for vector machines, but they have limited effectiveness in a parallel environment. Technical reasons for this are presented. The nature of automatic and nonautomatic code optimization is explored View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • On high-speed computing with a programmable linear array

    Publication Year: 1988, Page(s):425 - 432
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (800 KB)

    A simple programmable linear systolic array capable of solving a large number of problems drawn from a variety of applications is designed. The methodology is applicable to problems solvable by sequential algorithms that can be specified as nested FOT-loops of arbitrary depth. The algorithms of this form that can be computed on the array include 25 algorithms dealing with signal and image processi... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • The Convex C240 architecture

    Publication Year: 1988, Page(s):321 - 329
    Cited by:  Papers (4)  |  Patents (16)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (652 KB)

    A description is given of the C240, a tightly coupled, shared memory, parallel multiprocessor that supports up to 40-ns ECL/CMOS Cray-like processors. It is managed by a fully semaphored Unix operating system and can support up to 4 Gb of directly addressable physical memory. Convex proprietary compiler technology provides automatic vectorization and parallelization for Fortran, C and Ada. The all... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Compiling techniques for first-order linear recurrences on a vector computer

    Publication Year: 1988, Page(s):174 - 181
    Cited by:  Papers (4)  |  Patents (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (488 KB)

    The authors discuss high-performance methods for solving first-order linear recurrences on a vector computer, investigate automatic transformations, and develop the compiling techniques for first-order linear recurrence problems. The results show that the improved vector code generated by the vectorizing compiler run at the rate of 150 MFLOPS (million floating operations per second), in the case o... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Vectorized Monte Carlo molecular aerodynamics simulation of the Rayleigh problem

    Publication Year: 1988, Page(s):384 - 391
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (568 KB)

    A fully vectorized version of a one-dimensional Monte Carlo algorithm applied to molecular aerodynamics is presented. The algorithm is applied to the Rayleigh problem, an impulsively accelerated, heated flat plate, and is implemented on the CYBER 205. Issues relating to the details of vectorization, including the use of bit addressability, the maintaining of long vector lengths, and vector gather/... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Development of job-job step scheduler for NAL numerical simulator

    Publication Year: 1988, Page(s):200 - 206
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (324 KB)

    The authors present the concepts and functions of the job-job step scheduler that is the kernel of the software packages developed for the management of the NAL numerical simulator (NS). This scheduler is partially responsible for the high operating ratio of supercomputers in NS, and there is some guarantee of job turnaround View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A training program for scientific supercomputing users

    Publication Year: 1988, Page(s):342 - 349
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (840 KB)

    An 11-week training program aimed principally at graduate and postdoctoral students in computationally intensive fields is described. The program is designed to provide a balance between the specificity of computing center courses, the abstractness of computer science courses, and the personal contact of traditional apprentice approaches. It consists of seminars and clinics given by many visiting ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Mass storage support for supercomputing

    Publication Year: 1988, Page(s):217 - 221
    Cited by:  Papers (1)  |  Patents (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (336 KB)

    Mass storage support for supercomputing at Boeing Computer Services is satisfied by a locally developed storage system known as FMS, the File Management System. It runs on a dedicated back-end system and communicates with multiple processor types through several different local area networks. FMS follows the proposed IEEE mass storage reference model. The author discusses many of the problems and ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Vector and parallel processing of the nuclear reactor transient analysis code RELAP5

    Publication Year: 1988, Page(s):230 - 236
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (500 KB)

    An experiment on vector processing and multitasking of the nuclear reactor transient analysis code RELAPS has been carried out. Vector processing and multitasking were achieved by using the independency of the spatial meshes. The vectorization ratio is 83%. The performance ratio in the vector mode to that in the scalar mode is about three on the FACOM VP-100. For multitasking the spatial meshes ar... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A mass storage system for supercomputers based on Unix

    Publication Year: 1988, Page(s):279 - 286
    Cited by:  Papers (1)  |  Patents (3)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (580 KB)

    The authors present the design, implementation, and utilization of a large mass storage subsystem (MSS) for the numerical aerodynamics simulation. The MSS supports a large networked, multivendor Unix-based supercomputing facility. The MSS at Ames Research Center provides all processors on the numerical aerodynamics system processing network, from workstations to supercomputers, the ability to stor... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • The Joint Research Council's supercomputing unit

    Publication Year: 1988, Page(s):450 - 455
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (404 KB)

    A brief description is given of the Joint Research Councils' Supercomputer Unit, a Cray X-MP/48 installation for the use of academic researchers in the UK. Both hardware and system software are examined. Operational experience and scientific use of the machine are discussed View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Interactive scientific visualization and parallel display techniques

    Publication Year: 1988, Page(s):132 - 139
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (668 KB)

    A graphics environment for essentially real-time interactive scientific visualization of computational fluid mechanics is described. A researcher can interactively examine fluid data on a frame buffer with animated flow visualization diagnostics that mimic those in the experimental laboratory. This provides an effective and interactive way to analyze the underlying physical mechanisms and to compa... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • An evaluation of vector Fortran 200 generated by Cyber 205 and ETA-10 pre-compilation tools

    Publication Year: 1988, Page(s):106 - 113
    Cited by:  Papers (3)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (388 KB)

    Vectorizing precompilers such as KAP/205 and VAST-2 complement the efficient use of Fortran on the CDC Cyber 205. With the advent of the ETA-10 and its EOS/VSOS environment, the performance of these Fortran 200 preprocessors has come under closer scrutiny. The extent or quantity of vectorization that can be achieved has been examined elsewhere with reference to a test set of Fortran DO loops. The ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A Strassen-Newton algorithm for high-speed parallelizable matrix inversion

    Publication Year: 1988, Page(s):419 - 424
    Cited by:  Papers (6)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (384 KB)

    Techniques are described for computing matrix inverses by algorithms that are highly suited to massively parallel computation. The techniques are based on an algorithm suggested by V. Strassen (1969). Variations of this scheme use matrix Newton iterations and other methods to improve the numerical stability while at the same time preserving a very high level of parallelism. One-processor Cray-2 im... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • FACOM 6443 magneto-optic disk sub-system

    Publication Year: 1988, Page(s):266 - 271
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (368 KB)

    A rewritable optical disk drive is described. The drive has a large storage capacity of 8.9 Gb/spindle with a few seconds of access time. It can store data economically in a small floor space, and it can be used as a supplement or partial alternative to magnetic disk drives or tape library units. The authors discuss the drive technologies, support software, and application of this subsystem View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Some patterns of technological change in high-performance computers

    Publication Year: 1988, Page(s):312 - 320
    Cited by:  Papers (4)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (824 KB)

    The author describes the patterns of the unusually rapid change that high-performance computer technology is undergoing. He identifies the stages in which technological changes occur and then defines `change' as the first derivative of an information function that describes the state of a technology. He explores the driving forces that cause three generic patterns of technological change: incremen... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Compiling issues for supercomputers

    Publication Year: 1988, Page(s):164 - 173
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (612 KB)

    Two issues for supercomputer compilers are considered. Exact and inexact tests for computing data dependences and focus on the linearization scheme are first discussed. A program transformation call loop spreading is then presented that can be used to execute in parallel adjacent loops with interloop dependences View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Performance prediction for the Horizon supercomputer

    Publication Year: 1988, Page(s):48 - 50
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (176 KB)

    Horizon is a shared-memory multiple-instruction-stream-multiple-data-stream architecture currently under study. The performance of one Horizon processing element (PE) is quantified by user operations per instruction, the instructions per tick, and the basic clock rate. Assuming there is sufficient parallelism within a problem, the performance of one PE is multiplied by the number of PEs contained ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Elimination of bottlenecks in dynamic dataflow processors

    Publication Year: 1988, Page(s):80 - 87
    Cited by:  Papers (3)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (464 KB)

    A key component of a dynamic dataflow processor, the matching unit, has been identified as a major bottleneck. An alternative implementation for the matching unit is presented. This implementation increases the operating bandwidth of the unit by allowing token-matching operations to be done concurrently. The adoption of the proposed implementation also reveals throughput limitations of other units... View full abstract»

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