The performance-oriented design of linear storage hierarchies whtch are operating m muluprogramming environments ~s constdered An opt~mtzaUon model is superimposed upon an exponential queuing network model of the hterarchy, yteldmg a problem whose objectwe is to maxlmme throughput subject to a cost constraint.The dects~on variables are the speeds and capacmes of the various memory levels.It is shown that any local optimum is indeed a globally opumal solution to the problem.Several specml cases of and extensions to the basic problem are discussed, and some examples are given to dlustrate the usefulness and computational tractabihty of the problem KEY WORDS AND PHRASES linear storage hierarchies, miss ratio, multiprogrammmg, nonlinear programming, opttmization, performance-oriented design, queuing networks, storage technologies, systems modehng CR CATEGORIES 2 44, 4 32, 4.35, 4.6, 5 41, 5 5, 6 2, 6 34, 8 1, 8 3

article Free Access Share on Optimal Selection of CPU Speed, Device Capacities, and File Assignments Authors: Kishor S. Trivedi Department of Computer Science, Duke University, Durham, NC Department of Computer Science, Duke University, Durham, NCView Profile , Robert A. Wagner Department of Computer Science, Duke University, Durham, NC Department of Computer Science, Duke University, Durham, NCView Profile , Timothy M. Sigmon Department of Computer Science, Duke University, Durham, NC Department of Computer Science, Duke University, Durham, NCView Profile Authors Info & Claims Journal of the ACMVolume 27Issue 3July 1980 pp 457–473https://doi.org/10.1145/322203.322208Published:01 July 1980Publication History 44citation552DownloadsMetricsTotal Citations44Total Downloads552Last 12 Months35Last 6 weeks1 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my Alerts New Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteeReaderPDF

In this paper, a comparison of the performance of optimally designed computer systems with and without virtual memory is made. The computer systems in question are modeled by closed queuing networks of the central server type. The design of the systems is formulated as a nonlinear optimization problem where the objective function is to maximize the throughput subject to a nonlinear cost constraint. The decision variables are the speeds of the individual devices. This optimization problem is then solved by use of the Lagrange multiplier technique. The comparisons of the systems demonstrate the affect on performance of the addition of another I/O device to handle paging and the affect on performance of the additional overhead generated by the page fault handler. Also, the optimal amount of money to be spent on main memory is investigated.