Parallel and distributed programming is conceptually harder to undertake and to understand than sequential programming, because a programmer often has to manage the coexistence and coordination of multiple concurrent activities. The model of ‘Generative Communication ’ in Linda — a paradigm that has been developed for parallel computing — emphasizes the decoupling of cooperating parallel processes; thus, relieving the programmer from the burden of having to consider all process inter-relations explicitly. In many application areas, data is distributed over a multitude of heterogeneous, autonomous information systems. These systems are often isolated and an exchange of data among them is not easy. On the other hand, support for dynamic exchange of data is required to improve the business processes. Cooperative information systems enable such autonomous systems to interoperate. They are complex systems of systems which require a well designed and flexible software architecture. The Linda model had a great influence on research in parallel programming languages. Stimulated by this success, a Generative Communication Service, which offers a very flexible associative addressing mechanism based on metadata matching, has been developed for supporting interoperability of cooperative informationsystems. Some design patterns guided the construction of the resulting communication service that has been implemented on top of CORBA for an ODMG canonical data model.

Recently, much effort has been devoted to extending the shared memory paradigm to loosely-coupled machines that possess no physical shared memory. Two very different strategies for accomplishing this goal have emerged. The kernel-based approach attacks the problem from a low level. The language-based approach attacks the problem from a high level. Unfortunately, these two strategies do not currently converge. This paper describes a distributed operating system kernel, called ARCADE, which bridges the gap between the two strategies. ARCADE provides the low-level facilities needed to implement a variety of high-level distributed shared data paradigms. Furthermore, by providing an appropriate interface between applications and the kernel, ARCADE overcomes many of the problems encountered in earlier kernel-based implementations of distributed shared memory. 1. Introduction Shared memory can be an effective and efficient system structuring technique for uniprocessors and tightly-coupled mu...

Orca is a language for implementing parallel applications on distributed systems. This paper gives an overview of the Orca language and its underlying communication model, as well as the implementations and applications of the language. Also, it compares Orca with related systems, such as Linda and Shared Virtual Memory. 1. Introduction Orca is a language for implementing parallel applications on distributed systems. The language has been developed at the Vrije Universiteit in Amsterdam during the past five years. Orca is mainly being used for writing parallel programs that run on the Amoeba [1] distributed operating system, but in principle it is useful for any distributed system. This paper gives a brief overview of the Orca project. It discusses the Orca language and its underlying model, as well as the implementations and applications of the language. Finally, it compares Orca with several related systems, such as Linda and Shared Virtual Memory. The project is described in more d...

Simulator : : : : : : : : : : : : : : : : : : : : : : : : : 87 Appendix E. Examples of Human Performance Process Hierarchical Decomposition 92 Appendix F. Scalable Coherent Interfaces 96 Contents (continued) Chapter Page Appendix G. Synopses of Selected High Performance Parallel Machines 98 Appendix H. Glossary of Acronyms 102 References 105 List of Figures Figure Page 1.1 A Holarchy : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 17 2.1 Possible Paths for Human Performance Process Model Creation : : : : : : : 21 6.1 Numerical Aerodynamics Simulation Results for Embarassingly Parallel Benchmarks : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : : 40 6.2 CM2: Numerical Aerodynamics Simulation Benchmark Results : : : : : : : 41 6.3 Human Performance Process and Architectures : : : : : : : : : : : : : : : : 42 8.1 Heterogeneous Computing Environment : : : : : : : : : : : : : : : : : : : : 50 9.1 High Performance Systems Metrics : : :...