Abstract: Nowadays, Computing systems accessible to researchers with ”Grand Challenge” problems consist of a hardware mixture ranging from clusters of workstations to parallel supercomputers. This hardware is available via geographically distributed networks with various communication capabilities.

Abstract: Parallel and vector supercomputers are today considered basic research tools for several scientific and engineering disciplines. The novel architectural features of these computers which differ significantly from the von Neumann model are influencing the design and implementation of algorithms for numerical computation. Recent developments in large scale optimization take into account the architecture of the computer where the optimization algorithms are likely to be implemented. In the case of network optimization, in particular, we have seen significant progress in the design, analysis, and implementation of algorithms that are particularly well suited for parallel and vector architectures. As a result of these research activities, problems with several millions of variables can be solved routinely on parallel supercomputers. In this chapter, we discuss algorithms for parallel computing in

Abstract: Introduction to Client-Server Computing. Strategic Applications in Distributed Computing. Data Network Architectures. Building the Client-Server Architecture. Development of Client-Server Applications. Network Technology for Distributed Computing. Middleware and the Application Programming Interface. Industry-Supported Environments. Open Networking Standards (TCP/IP and OSI). Distributed Systems Management. Groupware -- Integrating People with Distributed Systems. Future Directions in Distributed Computing.

Abstract: Computational sciences require more computer power than current hardware technology is able to deliver at acceptable costs. Exploiting parallelism is the most promising way to obtain substantial performance increases. However, the absence of suitable parallel software delays a wider use of parallel systems. Automatic parallelization can help to overcome this problem. General automatic parallelization is a very difficult or even unsolvable problem. This paper shows that automatic parallelization becomes feasible when restrictions are imposed on the class of applications to be handled. The system presented in this paper is specialized for grid-based numerically intensive computations. An important feature of systems for parallel processing is scalability. Mechanisms to make the execution control of the automatic parallelization system scalable are presented. The validity of the concepts is shown by performance data. >

Abstract: The purpose of the paper is to introduce a frame-based approach to scientific computing on distributed computing platforms. The objective of the effort is to introduce well-defined mappings between the representations of physical phenomena as mathematical structures and the computational algorithms for modeling that phenomena on high-performance parallel computing platforms. Using this approach, we identify representations that solve classes of scientific problems based on the computational structure required for solution on parallel processors. Thus, once the representation is defined for a particular problem, the mapping onto a parallel computing environment can be managed automatically in software. We demonstrate the benefits of this approach by application to a combustion modeling problem. >

Abstract: For companies faced with the need to deliver new applications more quickly and at lower costs than before, client/server computing is becoming a reality. End-user computing, although not directly involved in creating the new applications, is affected by the shift in computing architecture.

Abstract: We explore briefly the historical notion of community, and indicate how technology is changing our view of what constitutes a community. We then examine the kinds of value (information, communication and service) that can be and “ought to” be delivered over computer networks, and suggest adaptations of current network structures that could deliver these values effectively and efficiently. The focus will be on the local geographical community, and on the nature of interconnections between local communities. Finally, we consider technological and social challenges inherent in providing community computing and networking. THE INDIVIDUAL’S WORLD VIEW Whatever else can be said about the universe, it is not absolute. Each of us is, in our own view, the center of it; thus it has one cente~ it has many centers; it has no center. The way you and I perceive the universe, or any part of it, will depend on our unique perspectives. We do not necessarily value the same things, and the value of a thing to us depends upon how it relates to us as individuals. In the highest sense, we would like to believe that Technology exists to serve people in the pursuit of higher human values such as knowledge, creativity, community, culture, integrity and wisdom. Technology is not an end in itself, but serves to empower, to connect, to enhance, and to assist. Even in that context, however, the role of technology is subject to personal interpretation. The value of the technology to an individual depends on that individual’s personal goals. Such goals may be subsumed in the above list, or may be more immediate goals such as to save effort, save time, make money, exert influence, or cultivate friendships. From our individual perspective, the world can be viewed as a large onion, with the self at the core. Subsequent layers might correspond to family, neighborhood, community, area, state, region, nation, and finally world. In a narrow sense, one might think of individuals interacting within alfamily; families interacting within a neighborhood; neighborhoods interacting within a community, and so on. In practice, of course, the various kinds of interactions are far more complex than this; indeed, from our individual perspective, it is we, as individuals, who interact with various amalgamations of other individuals. WHAT IS COMMUNITY? When we hear of a neighborhood, many of us automatically think of some collection of homes or other living units in geographical proximity. Indeed, that is essentially the dictionary definition of neighborhood. However, technology and Fred Rogers began a redefinition of neighborhood more than a quarter century ago, and the term has come to include mental/spiritual proximity in addition to geographical proximity. Similarly, “Community” at one time tended to connote geographical proximity (city or town), although it has also been used to indicate some more tightly-focused commonality of interest as well. We are used to hearing about a “University Community” in conjunction with an institution of higher education or a “Community of Faith” in conjunction with a church. Historically, neighborhood and community have been defined in terms of geographical boundaries, technological boundaries (such as local telephone calling area), and political boundaries. More recently, new forms of community have been defined by social boundaries, technological boundaries, and experiential boundaries. What characterizes community in the modern world, then, is lmerely some form of interdependence: perhaps a common geographic location, or common interests, or common beliefs, or common culture, or common experience. TECHNOLOGY AND COMMUNITY Technology has fostered new forms of community by dramatically diminishing the importance of the old boundI Permission to copy without fee all or part of this material is granted provided that the copies are not made or distributed for direct commercial advantage, the ACM copyright notice and the title of the publication and its akzta appea~ and notice is given that copying is by permission of the Association for Computing Machinery, to copy otherwise, or to republish, requires a fee anoYor specljic permission. 01994 ACM ISBN 0-89791-656-5/94/459940 $3.50 Meet the Shadowy Future 35 aries. Television has for years had the capacity to bring world events into our personal space in a timely and dramatic manner. However, that does not, in general, inject our person into the world space, and so does not involve us personally in a community. On the other hand, computers and particularly computer networking have had the opposite effect; they let us as individuals explore the world space and (more or less frequently) “meet” other people. This tends to make us as individuals feel we are a part of a community (whether or not we actually are). In any case, technological networking has dramatically decreased the importance of geographical proximity and other factors that used to be definitive of community, and has fostered the development of geographically and politically diverse communities formed around common interests, common beliefs, or common experiences. The Computing Community is at once the parent and the child of modem computing technology. WHAT’S OUT THERE? At times, the discussion of the marvels of “the Internet” seems to be all hyperbole and no substance. Everyone knows “the Intemet” is a Good Thing, but comparatively few can provide ready examples of benefits that have accrued directly from the existence of computer networking. Many of the arguments for acquiring new technology seem to follow the Field of Dreams philosophy, “if you build it, they will come.” In the case of technology, however, “they” sometimes seems to refer to nothing more than bucketful of bits. Clearly, none of the technology is free. Equipment costs money; transportation of data costs money; and all of it occupies a great deal of effort on the part of many different people (which, after all, money was invented to represent). Somebody has to pay for all this; and that somebody is clearly the person at the center of the universe. Let us consider, then, what resources in the various layers of the universal onion might be valuable to an individual. Personal and family resources tend to be “my own” private resources. Personal computers and software, if it’s perceived that this will somehow advance our progress toward our goals. If we make that investment, most likely we have some personal information that we store on our personal computer; such things as financial records, an inventory of our “stuff,” family history, our embryonic bestselling novel, and so on. We may buy an encyclopedia on CD-ROM, baseball statistics, or other information of interest to us. If we see good reason, we may invest in a telephone line and a modem so we can get beyond our personal space. In all these cases, we justify these expenses to ourselves, in conjunction with our other priorities. In most cases, unlike the ones to follow, information stored on personal or family computers is not intended for public access. 36 Neighborhood and community resources are available on many college and university campuses through the Campuswide Information Systems. These systems typically include such items as telephone directories, events calendars, policies and procedures, university catalogs, and other information for delivery, although some include facilities for applying for admission or getting a computer account. At present, neighborhood and community resources are available by network in only a few cities and towns, but the number is growing rapidly. Things such as local government services, community calendars, and telephone books are only a few of the features that could be included in such a system. The pattern of these two layers continues throughout the remaining layers of the universal onion. Each layer may add its own value (services) to the network, and one of the values each layer may add is access to other layers of other onions. So, for example, in an regional computing system, one might find access to other state computing systems within the region, in addition to access to information on regional resources such as the National Center for Supercomputing Applications at the University of Illinois. Examples are not difficult to come by, and we will not belabor them here. Most resources at the outer layers of the onion are things that are, or have historically been, available in other forms. Telephone directories, for instance, are routinely printed and distributed for most communities (geographical and otherwise). The cost of making these resources available via network must be borne by someone, and the benefit of doing so must justify the cost. Either the individual seeking the information will pay, or the organization supplying it will. The information or service will continue to be a viable part of the network only so long as someone benefits enough to continue footing the bill. WHAT ISA NETWORK? We need to distinguish between the structural network and the conceptual network. Structurally, a network consists of things to be connected together (“nodes”) along with cables and electronics necessary to move signals from one node to another. In our onion model, the self can be networked to the self by a standalone computer (these are of decreasing size and increasing portability, so we will not discuss the obvious counterexample of home vs. oi%ce). In many families, a single standalone computer is adequate, although some families are connecting multiple computers together via local area networks. Connections from the family environment to other layers tend to be via telephone line and modem. While this can typically connect one with any layer, the advantages of connecting to a resource within the local calling area are, in most cases, obvious.

Abstract: For the application of high-performance, distributed computing systems in solving complex computational problems with high real-time demands and in order to meet the real-time requirements on the underlying communication system the paper deals with application fundamentals of distributed, high-performance computing and networking systems. Evaluation of hard-bounded, worst-case, real-time computing and networking performances and latency and bandwidth evaluation of distributed, high-performance systems shows that such systems have forced us to deal with the propagation delay within the system due to the finite speed of light. The user must pay attention to his file sizes and how latency will affect his applications. At the application level, it is important to use parallelism.

Abstract: This paper hints at models and mechanisms which are part of current distributed systems research, and which may be of interest in the area of distributed, parallel computing as well. In this context, an experimental toolkit which allows for object-oriented programming of distributed, failure-resilient applications is presented. The toolkit, called Electra, supports novel features like object-oriented communication, object-groups, and reliable multicast. We will compare the performance of a compute intensive application implemented on Electra, on PVM, on a transputer, and on two different Linda systems.

Abstract: A network of supercomputers and high-performance workstations appears to be the only reasonable way to provide adequate computing resources for the Grand Challenge problems of the next century. Such a collection of computers and supporting software environments is called a virtual computing environment (VCE). The paper describes the motivation and goals of the VCE project, followed by a description of the system. The paper concentrates on the runtime aspects of the VCE, and concludes with a discussion of a small prototype system that has been built using the Isis distributed toolkit.<>

Abstract: The general area of secure distributed computing and the interplay between distributed computing and security/ cryptography research is reviewed. Recent theoretical and practical developments are discussed.

Abstract: The emerging generation of complex software systems presents significant challenges that must be addressed with new development technologies. Such systems are highly distributed and employ many heterogeneous processors, some of which may be parallel processors. Additionally, there are many non-functional requirements (related to timing, reliability, security and fault tolerance).

Abstract: We report on a portable, easily usable communication environment, ‘Sciddle’, for distributing computations over heterogenous networks of UNIX computers.

Abstract: The efficient parallel execution of grid-oriented scientific calculations requires the partitioning of the grid in such a way that the work load is equally distributed over the processors of the parallel machine and that communication among the processors is minimized. For unstructured static grids, good partitions are obtained by applying the recursive spectral bisection heuristic to the interdependency graph of the grid. We describe how even in case of dynamically changing grids, grid-oriented problems can be formulated as graph partitioning problems for the purpose of load balancing. We further describe an alternative spectral bisection algorithm that yields better partitions than the standard algorithm for the graphs that model dynamic load balancing problems. >

Abstract: This paper aims at illustrating a methodology of generation of an altimetrical data base for mobile radio applications and a procedure developed to manage it, using a GIS (Geographic Information System). The system uses as input a set of digitized maps in the form of contour lines and elevation points obtained from cartographic maps at different scales (1:250,000, 1:50,000 and 1:10,000), and generates a final data base in the form of a set of regular matrices of altimetrical values with an appropriate grid size. The operator is allowed to check at each step the state of the maps under development and, if necessary, to correct errors or to change parameters. After a brief introduction on the principles of GISs, it will be shown how their functions of inserting, storing and managing altimetrical data have been used to create a data base useful for electromagnetic computations and how a special user interface allows the operator to select an area on the territory and automatically generate the corresponding altimetrical matrix with a suitable grid size. Finally, a few applications are described, as computer aided tools for cellular land mobile radio networks planning, concerning documentation of radio coverages, radio link analyses, intervisibility analyses and 3-D representations, used both for technical and commercial purposes.

Abstract: Distributed systems with point-to-point interconnection networks are natural candidates for real-time fault-tolerant communication because parallel processing and communication as well as fault-tolerance can be achieved using multiple processors and interconnection paths between every pair of nodes. However, due to the contention among randomly-arriving messages at each node/link and multi-hops between the source and destination that a message must travel, it is difficult to guarantee the timely delivery of the messages. The goal of this chapter is to remove this difficulty and simultaneously realize the potential of distributed systems for high performance and high reliability.