PSEE: A tool for parallel systems learning

E. Luque, R. Suppi, J. Sorribes, E. Cesar, J. Falguera, M. Serrano

    Research output: Contribution to journalArticleResearchpeer-review

    4 Citations (Scopus)

    Abstract

    Programs for parallel computers of distributed memory are difficult to write, understand, evaluate and debug. The design and performance evaluation of algorithms is much more complex than the conventional sequential one. The technical know-how necessary for the implementation of parallel systems is already available, but a critical problem is in the handling of complexity. In parallel distributed memory systems the performance is highly influenced by factors as interconnection scheme, granularity, computing/communication ratio, algorithm topology, parallel languages, and operating system policies. Interaction between these factors is not easy to understand and often unpredictable. This paper presents the PSEE (Parallel System Evaluation Environment), an interactive graphical environment, which permits to study the behaviour of parallel distributed memory systems, as well as several enhancements to increase the goodness of this tool. PSEE is an easy-to-use environment that enables parallel distributed memory systems programmers take decisions about the behaviour program and parallel computer in terms such as scalability, tuning and performance of the underlying parallel machine.
    Original languageEnglish
    Pages (from-to)319-339
    JournalComputers and Artificial Intelligence
    Volume15
    Issue number4
    Publication statusPublished - 1 Dec 1996

    Keywords

    • Parallel processor simulation
    • Parallel programming
    • Performance visualization
    • Simulation environment

    Fingerprint Dive into the research topics of 'PSEE: A tool for parallel systems learning'. Together they form a unique fingerprint.

    Cite this