By Topic

IBM Journal of Research and Development

Issue 5 • Date Sep. 1975

Filter Results

Displaying Results 1 - 12 of 12
  • Preface

    Page(s): 434
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (181 KB)  

    In a computer system there is a wide spectrum of actions that are to be performed simultaneously in order to maximize throughput and minimize the response time of the system. The diversity of action and reaction which takes place in a large computer system presents serious problems to anyone trying to encompass that entire spectrum. No single theory of optimization-finding the best way to carry out an action to produce a specific result-embraces all the important aspects because each domain of application has unique characteristics that determine the most appropriate approach. There are many problems of long-standing interest which have eluded solution, partly because the problems have not been clearly defined and partly because the most significant parameters to be used in tractable models have not been identified. The eight papers in this issue introduce some important and novel concepts in the analysis and design of computer systems. New viewpoints have been developed for the analysis of significant aspects such as multiprocessing, multiprogramming, system communication, storage management, scheduling, and validation of simulation models. The concepts presented in this group of papers contain numerous warnings and restrictions, which serve to call attention to those aspects of the discussion that must be modified in order to represent in a more realistic manner the physical processes of existing systems. Success in the application of the described models depends on careful formulation by the user and an understanding of the assumptioas and restrictions involved in their use. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Computation of Lower Bounds for Multiprocessor Schedules

    Page(s): 435 - 444
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (552 KB)  

    A multiprocessing system composed of identical units is considered. This system is executing a set of partially ordered tasks, with known execution times, using a non-preemptive scheduling strategy. Lower bounds on the number of processors required to compute the tasks before a deadline, and on the minimum time to execute the tasks with a fixed number of processors, are of great value for the determination of the corresponding optimal schedules. In this paper, methods for the efficient computation of the lower bounds obtained by Fernández and Bussell are discussed. Computational improvements for the case of general partial orders are reported, and further reductions of the number of operations are shown to be possible for special graphs (trees, independent chains, independent tasks). View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Dynamic Partitioning of the Main Memory Using the Working Set Concept

    Page(s): 445 - 450
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (419 KB)  

    An algorithm to divide the main memory among N competing programs with different characteristics, running in a multiprogramming and virtual memory environment, is proposed. The algorithm is based on an optimal allocation policy, which is derived in this paper, using the concept of the working set. A brief description of the hardware implementation of the algorithm is also presented. It is shown that under this optimal allocation policy “the value of a page-frame” (the amount of reduction in the page fault rate if an additional page frame is allocated to that program) to each program is the same. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Study of Memory Partitioning for Multiprogramming Systems with Virtual Memory

    Page(s): 451 - 457
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (421 KB)  

    In this paper, we investigate the effect that the shape of the lifetime function has on the optimal partition of the main memory of a computer among N programs, where the criterion of optimality is maximization of CPU utilization. We used a simple queuing model as a base for understanding this interrelationship. The lifetime function is the average of the execution intervals of a program as a function of the amount of memory allocated. When the lifetime function is convex and is proportional to mα, where m is the size of memory, then the optimal partition is obtained by dividing the main memory equally among q of the N programs (q is the optimal degree of multiprogramming). Thus, the best partition is always one of two policies; allocate all memory equally among the q programs or allocate all memory to one program. When the lifetime function has a degenerate S shape (is proportional to mα when m ≤ m0 and remains constant beyond m0), then there exists a memory size m such that no program can have a memory other than m or m0 if any program has a memory size greater than m0 each other program should have a memory size that is equal to or greater than m0 View full abstract»

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

    Page(s): 458 - 462
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (410 KB)  

    A recently developed method for estimating confidence intervals when simulating stochastic systems having a regenerative structure is reviewed. The paper is basically tutorial, but also considers the pragmatic issue of the simulation duration required to obtain valid estimates. The method is illustrated in terms of simulating the M/G/1 queue. Analytic results for the M/G/1 queue are used to determine the validity of the simulation results. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Regenerative Simulation of a Queuing Model of an Automated Tape Library

    Page(s): 463 - 475
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (641 KB)  

    Recently, techniques have been developed for estimating confidence intervals when simulating stochastic systems having a regenerative structure. These techniques are applied to the simulation of a queuing model of a computer system's automated tape library. Theoretical and practical issues related to the application of these techniques are addressed. An interesting feature of the automated tape library represented in the queuing model is that certain queues have finite capacity; when these queues are filled to capacity certain services are prevented from occurring. The regenerative techniques are used in conjunction with multiple comparison procedures to make statistically valid statements about the effect of the finite queue capacities on performance. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Sequential Server Queues for Computer Communication System Analysis

    Page(s): 476 - 485
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (566 KB)  

    A queuing model with two sequential servers is developed to analyze performance in computer and communication systems. In one case the CPU is the first server and the terminal and its associated communications equipment are the second server. In a second case the CPU and the channel are the first server and the auxiliary storage device is the second server. We study the queuing behavior of the sequential server systems with Poisson arrivals, general service time distributions, and several service disciplines, including bulk arrivals, message priorities, and the input and output queues. The stationary distributions of the queue lengths and waiting times are determined by using an imbedded Markov chain analysis. Several examples are given to illustrate the applications of these models to practical problems. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Analysis of a Loop Transmission System with Round-Robin Scheduling of Services

    Page(s): 486 - 493
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (484 KB)  

    A finite population, multi-queue model is developed for a loop transmission system. Approximate expressions for the state transition matrix and other system variables are derived in recursive forms. It is also shown that a number of useful system parameters, such as average message response time, average cycle time, and average response time conditioned on message length, can be obtained. The analytical results have been validated by simulation. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Optimal Scheduling Strategies for Real-Time Computers

    Page(s): 494 - 504
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (883 KB)  

    In order to fulfill response time constraints in real-time systems, demands are often handled by means of sophisticated scheduling strategies. This paper first shows how to describe and analyze arbitrary combinations of preemptive and non-preemptive (head-of-the-line) priority strategies and, second, presents an algorithm that yields the optimal priority strategy, taking into consideration constraints on the response time. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Recent Papers by IBM Authors

    Page(s): 505 - 507
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (335 KB)  

    Reprints of the papers listed here may usually be obtained by writing directly to the authors. The author's IBM divisions are identified as follows: DPD is the Data Processing Division; FED, Field Engineering Division; FSD, Federal Systems Division: GPD, General Products Division; GSD, General Systems Division; OPD, Office Products Division; RES, Research Division; SCD, System Communications Division; and SPD, System Products Division. Papers are listed alphabetically by name of journal. View full abstract»

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

    Page(s): 508 - 510
    Save to Project icon | PDF file iconPDF (230 KB)  
    Freely Available from IEEE
  • Authors

    Page(s): 511 - 512
    Save to Project icon | PDF file iconPDF (256 KB)  
    Freely Available from IEEE

Aims & Scope

The IBM Journal of Research and Development is a peer-reviewed technical journal, published bimonthly, which features the work of authors in the science, technology and engineering of information systems.

Full Aims & Scope

Meet Our Editors

Editor-in-Chief
Clifford A. Pickover
IBM T. J. Watson Research Center