This paper is concerned with the performance analysis of a deadlock free protocol for limited cap... more This paper is concerned with the performance analysis of a deadlock free protocol for limited capacity ring networks. There are ¡ nodes on the ring. Each node is connected with an emitter and a receiver of messages and is equiped with a buffer of capacity two for receiving and transmitting messages. The ring is of uni-direction so that every node has a communication predecessor and a successor.
Abstract: This paper is concerned with the performance analysis of a deadlock free protocol for l... more Abstract: This paper is concerned with the performance analysis of a deadlock free protocol for limited capacity ring networks. There are,nodes on the ring. Each node is connected with an emitter and a receiver of messages,and is equiped with a buffer of capacity two for receiving and transmitting mess ages. The ring is of uni‐direction so that every node has
The Sloop project, launched in 1995, works on three main domains: (1) discrete event simulation, ... more The Sloop project, launched in 1995, works on three main domains: (1) discrete event simulation, (2) parallel object-oriented languages and (3) interconnection networks. One of the research objective of the project is the development of Prosit, a sequential and distributed object-oriented workbench for discrete event simulation. Object orientation is used to ensure two main concerns: -- performance on conventional, networked and multi-processor machines -- versatility and ease of use. We have now achieved the rst part of Prosit. We have completely specied the simulation and the execution models, and we have a sequential C++ version of the simulator. This report concentrates on the modelling process and the object oriented architecture of the sequential simulator. As an illustration of possible applications, we present the classes supporting the queueing network paradigm.
In this paper we outline the technical choices that we did in order to distribute the Hipertrans ... more In this paper we outline the technical choices that we did in order to distribute the Hipertrans traffic simulator: to use a spatial distribution of the model and a time-step synchronisation. In order to do a good spatial distribution we developed a partitioning module. This module uses two heuristics: Simulated Annealing and Genetic Algorithms for producing a nearly optimal partitioning of the traffic network.
The authors define and evaluate two policies (NA-policy, A-policy) for parallel execution of prog... more The authors define and evaluate two policies (NA-policy, A-policy) for parallel execution of program tree structures. Via a probabilistic model they analytically determine, for each policy, the Laplace-Stieltjes transform for the tree processing time distribution. The acceleration of the program execution time achieved when adding processors to a single processor environment, is computed and plotted for each policy.
ABSTRACT An efficient way of building a polyhedral approximation of a set of points in 3-D space ... more ABSTRACT An efficient way of building a polyhedral approximation of a set of points in 3-D space is described. The points are the vertices of a planar graph embedded in a surface of genus 0 and are obtained by a laser range finder. The technique presented here is a generalization of an existing algorithm (R. Duda and P. Hart, Pattern Classification and Scene Analysis, Wiley-Interscience, New York 1973) for the polygonal approximation of a simple curve in 2-D space.
2nd European School of Computer Science, Parallel Programming Environments for High Performance Computing ESPPE’96, 1996
1 Introduction We give an overview of the SLOOP project. A synthetic view of its organization, an... more 1 Introduction We give an overview of the SLOOP project. A synthetic view of its organization, and each component are presented. This presentation however, primarily focuses on the parallel object-oriented language and simulation aspects. 2 The SLOOP project Our group (INRIA{I3S/CNRS{Univ. of Nice Sophia Antipolis) works in three main areas: ... 1. parallel and distributed discrete event simulations, ... These three different research goals articulate one with another in the construction of the Sloop system: each level uses the primitives and ...
The Prosit project, launched in 1991, is devoted to thedevelopment of a new discrete event simula... more The Prosit project, launched in 1991, is devoted to thedevelopment of a new discrete event simulation systembased on the object paradigm. Object orientation is usedto ensure two main concerns:ffl performance on both conventional and multiprocessormachinesffl versatility and ease of use.In particular, Prosit user programs (models and experimentdescriptions) should need only a new compilationso as to run efficiently on either a
... issue we will not comment on it, but rather say that duplication of the customer profile on t... more ... issue we will not comment on it, but rather say that duplication of the customer profile on the ... that problem, even though it implies a communication overhead to maintain consistency among all copies of the profile. ... International computer science institute, UC Berkeley, June 1991. ...
This paper outlines a draft definition for EC++. EC++ is a framework within which parallel C++ sy... more This paper outlines a draft definition for EC++. EC++ is a framework within which parallel C++ systems can be standardised. The framework will add portability to parallel C++ systems and will run across a variety of hardware architectures, while encompassing as wide a set of parallel computing models and paradigms as possible, both standard models and user extensible models. This is done entirely within standard C++, i.e. without syntactic extensions to C++.
This paper shows how the Prosit system, a new C + +-based framework for both sequential and distr... more This paper shows how the Prosit system, a new C + +-based framework for both sequential and distributed discrete event simulation, developed at INRIA-Sophia-Antipolis, makes an easy and efficient integration of classical discrete event simulation and of new high speed simulation techniques based on evolution equations possible. This demonstrates in particular the feasibility of distributed simulations involving simulators of different nature. Important applications of these techniques may be found in the simulation of communication switches, as illustrated in an example.
: This report describes PARSEVAL (PARalellizing Simulations for performanceEVALuation), an experi... more : This report describes PARSEVAL (PARalellizing Simulations for performanceEVALuation), an experimental distributed simulation workbench dedicatedto queueing systems and running on a network of Transputers attached to a Sunworkstation. It uses an implementation of the conservative method proposed byChandy and Misra for the synchronization of processes.After a brief description of the structure of PARSEVAL, the different parts thatmake it up and
... The Prosit project is devoted to the develop-ment of a new discrete event simulation system. ... more ... The Prosit project is devoted to the develop-ment of a new discrete event simulation system. ... di erent simulated component there must be a class embodying the corresponding functional behavior and implementing the desired services. ... 1. The class already exists in the library. ...
This paper presents the development of a parallel object-oriented language which is an extension ... more This paper presents the development of a parallel object-oriented language which is an extension of C++, called C ++//. C ++// offers reusability, flexibility and extensibility in concurrent programming through a set of language primitives (indeed a Meta-Object Protocol), independent of any parallel paradigm. It permits us to build libraries of nearly all concurrent programming models. One of them, presented here, is an MIMD model based on data-flow synchronizations (wait-by-necessity). C++// runtime main concern is versatility so its implementation is in fact an interface to any low-level runtime support.
This paper is concerned with the performance analysis of a deadlock free protocol for limited cap... more This paper is concerned with the performance analysis of a deadlock free protocol for limited capacity ring networks. There are ¡ nodes on the ring. Each node is connected with an emitter and a receiver of messages and is equiped with a buffer of capacity two for receiving and transmitting messages. The ring is of uni-direction so that every node has a communication predecessor and a successor.
Abstract: This paper is concerned with the performance analysis of a deadlock free protocol for l... more Abstract: This paper is concerned with the performance analysis of a deadlock free protocol for limited capacity ring networks. There are,nodes on the ring. Each node is connected with an emitter and a receiver of messages,and is equiped with a buffer of capacity two for receiving and transmitting mess ages. The ring is of uni‐direction so that every node has
The Sloop project, launched in 1995, works on three main domains: (1) discrete event simulation, ... more The Sloop project, launched in 1995, works on three main domains: (1) discrete event simulation, (2) parallel object-oriented languages and (3) interconnection networks. One of the research objective of the project is the development of Prosit, a sequential and distributed object-oriented workbench for discrete event simulation. Object orientation is used to ensure two main concerns: -- performance on conventional, networked and multi-processor machines -- versatility and ease of use. We have now achieved the rst part of Prosit. We have completely specied the simulation and the execution models, and we have a sequential C++ version of the simulator. This report concentrates on the modelling process and the object oriented architecture of the sequential simulator. As an illustration of possible applications, we present the classes supporting the queueing network paradigm.
In this paper we outline the technical choices that we did in order to distribute the Hipertrans ... more In this paper we outline the technical choices that we did in order to distribute the Hipertrans traffic simulator: to use a spatial distribution of the model and a time-step synchronisation. In order to do a good spatial distribution we developed a partitioning module. This module uses two heuristics: Simulated Annealing and Genetic Algorithms for producing a nearly optimal partitioning of the traffic network.
The authors define and evaluate two policies (NA-policy, A-policy) for parallel execution of prog... more The authors define and evaluate two policies (NA-policy, A-policy) for parallel execution of program tree structures. Via a probabilistic model they analytically determine, for each policy, the Laplace-Stieltjes transform for the tree processing time distribution. The acceleration of the program execution time achieved when adding processors to a single processor environment, is computed and plotted for each policy.
ABSTRACT An efficient way of building a polyhedral approximation of a set of points in 3-D space ... more ABSTRACT An efficient way of building a polyhedral approximation of a set of points in 3-D space is described. The points are the vertices of a planar graph embedded in a surface of genus 0 and are obtained by a laser range finder. The technique presented here is a generalization of an existing algorithm (R. Duda and P. Hart, Pattern Classification and Scene Analysis, Wiley-Interscience, New York 1973) for the polygonal approximation of a simple curve in 2-D space.
2nd European School of Computer Science, Parallel Programming Environments for High Performance Computing ESPPE’96, 1996
1 Introduction We give an overview of the SLOOP project. A synthetic view of its organization, an... more 1 Introduction We give an overview of the SLOOP project. A synthetic view of its organization, and each component are presented. This presentation however, primarily focuses on the parallel object-oriented language and simulation aspects. 2 The SLOOP project Our group (INRIA{I3S/CNRS{Univ. of Nice Sophia Antipolis) works in three main areas: ... 1. parallel and distributed discrete event simulations, ... These three different research goals articulate one with another in the construction of the Sloop system: each level uses the primitives and ...
The Prosit project, launched in 1991, is devoted to thedevelopment of a new discrete event simula... more The Prosit project, launched in 1991, is devoted to thedevelopment of a new discrete event simulation systembased on the object paradigm. Object orientation is usedto ensure two main concerns:ffl performance on both conventional and multiprocessormachinesffl versatility and ease of use.In particular, Prosit user programs (models and experimentdescriptions) should need only a new compilationso as to run efficiently on either a
... issue we will not comment on it, but rather say that duplication of the customer profile on t... more ... issue we will not comment on it, but rather say that duplication of the customer profile on the ... that problem, even though it implies a communication overhead to maintain consistency among all copies of the profile. ... International computer science institute, UC Berkeley, June 1991. ...
This paper outlines a draft definition for EC++. EC++ is a framework within which parallel C++ sy... more This paper outlines a draft definition for EC++. EC++ is a framework within which parallel C++ systems can be standardised. The framework will add portability to parallel C++ systems and will run across a variety of hardware architectures, while encompassing as wide a set of parallel computing models and paradigms as possible, both standard models and user extensible models. This is done entirely within standard C++, i.e. without syntactic extensions to C++.
This paper shows how the Prosit system, a new C + +-based framework for both sequential and distr... more This paper shows how the Prosit system, a new C + +-based framework for both sequential and distributed discrete event simulation, developed at INRIA-Sophia-Antipolis, makes an easy and efficient integration of classical discrete event simulation and of new high speed simulation techniques based on evolution equations possible. This demonstrates in particular the feasibility of distributed simulations involving simulators of different nature. Important applications of these techniques may be found in the simulation of communication switches, as illustrated in an example.
: This report describes PARSEVAL (PARalellizing Simulations for performanceEVALuation), an experi... more : This report describes PARSEVAL (PARalellizing Simulations for performanceEVALuation), an experimental distributed simulation workbench dedicatedto queueing systems and running on a network of Transputers attached to a Sunworkstation. It uses an implementation of the conservative method proposed byChandy and Misra for the synchronization of processes.After a brief description of the structure of PARSEVAL, the different parts thatmake it up and
... The Prosit project is devoted to the develop-ment of a new discrete event simulation system. ... more ... The Prosit project is devoted to the develop-ment of a new discrete event simulation system. ... di erent simulated component there must be a class embodying the corresponding functional behavior and implementing the desired services. ... 1. The class already exists in the library. ...
This paper presents the development of a parallel object-oriented language which is an extension ... more This paper presents the development of a parallel object-oriented language which is an extension of C++, called C ++//. C ++// offers reusability, flexibility and extensibility in concurrent programming through a set of language primitives (indeed a Meta-Object Protocol), independent of any parallel paradigm. It permits us to build libraries of nearly all concurrent programming models. One of them, presented here, is an MIMD model based on data-flow synchronizations (wait-by-necessity). C++// runtime main concern is versatility so its implementation is in fact an interface to any low-level runtime support.
Uploads
Papers by Philippe Mussi