Many engineers are still reluctant to adopt advanced object-oriented technologies (such as high modularity, dynamic binding, automatic garbage collection, etc.) for embedded systems with real-time constraints, because of their supposed inefficiency. We set ourselves into the context of building telecommunication systems with a standard object-oriented analysis and design approach. We describe how we use relevant design patterns, followed with an implementation in a pure object-oriented language (Eiffel) to conciliate the needed efficiency with the benefits of the object-oriented approach —flexibility, dynamic configurability, maintenability, portability, etc. We discuss a case study based on the implementation of SMDS (Switched Multi-megabits Data Service) servers featuring high-throughput and low-delay transmissions and respecting the real-time constraints of SMDS.

: We describe in this paper a Parallel Observable virtual Machine (POM), which provides a homogeneous interface upon the communication kernels of parallel architectures. POM was designed so as to be ported easily and efficiently on numerous parallel platforms. It provides sophisticated features for observing distributed executions. Key-words: Distributed memory parallel computers, virtual machine, communication library, observation, traces (R'esum'e : tsvp) guidec@irisa.fr maheo@irisa.fr Unite de recherche INRIA Rennes IRISA, Campus universitaire de Beaulieu, 35042 RENNES Cedex (France) Telephone : (33) 99 84 71 00 -- Telecopie : (33) 99 84 71 POM : une machine parall`ele virtuelle incorporant des m'ecanismes d'observation R'esum'e : Nous d'ecrivons dans cet article une machine parall`ele virtuelle observable, la POM. Celle-ci offre une interface homog`ene au dessus des syst`emes de communication des architectures parall`eles. Elle a 'et'e con¸cue en vue d'un portage ais'e et ...

: Encapsulating parallelism and synchronization code within object-oriented software components is a promising avenue towards mastering the complexity of the distributed memory supercomputer programming. However, in trying to give application programmers benefit of supercomputer power, the library designer generally resorts to low level parallel constructs, a time consuming and error prone process. To solve this problem we introduce a new abstraction called Parallel Operators. A Parallel Operator exists simultaneously on all processors involved in a distributed computation: it acts as a single entity capable of processing shared or distributed data in parallel. In this paper we reify this concept in our Eiffel Parallel Execution Environment (EPEE) context, and we show that it is both natural and efficient to express computations over large shared or distributed data structures using Parallel Operators. We illustrate our approach with an Parallel Operator based solution for the well-kn...

Parallel computers offer the ability to reduce the run-time of large and complex discrete event simulations. Most research into Parallel Discrete Event Simulation (PDES) is based on libraries for existing languages, such as C and C++. This library-based approach exposes the user to the synchronisation protocols used to ensure that events are processed in the correct order, and produces simulators that are less expressive than simulation-oriented languages (for example, Simula and MODSIM II). The use of a simulation-oriented language for PDES allows the synchronisation protocols to be hidden from the user and enables optimisations to be performed that are usually difficult or cumbersome, such as granularity control. A simulation-oriented language could therefore improve the usability of PDES, the lack of which has generally inhibited its widespread take-up. This thesis focuses on the run-time support needed for simulation-oriented languages for PDES using a new language cal...

Control and data parallelism are two complementary but often mutually exclusive paradigms used to program massively parallel systems. We propose to encapsulate both control and data parallelism in regular classes of a sequential Object Oriented language: a SPMD programming model is used and thus no language extensions are needed, provided a Shared Virtual Memory is available. We show how these ideas are implemented in EPEE, our Eiffel Parallel Execution Environment. As an example, we present the implementation of both paradigms on a toy linear algebra example and show how they can interoperate. We conclude with some performance results and prospectives remarks. 1 Introduction High Performance Fortran (HPF) is a new release of Fortran for programming massively parallel architectures. It is based on the so called data parallelism model where the set of data involved in a computation is split into partitions, to which processes are associated: this makes it possible to benefit from one ...