Emmanuel Melin

FlowVR: A Middleware for Large Scale Virtual Reality Applications

This paper introduces FlowVR, a middleware dedicated to virtual reality applications distributed on clusters or grid environments. FlowVR supports coupling of heterogeneous parallel codes and is component oriented to favor code reuse. While classical communication paradigms focus on either a synchronous approach (FIFO channels) or an asynchronous one (sampling), FlowVR enables a large range of intermediate policies to better balance the application performance between levels of details, latencies and refresh rates.

Net Juggler: running VR Juggler with multiple displays on a commodity component cluster

Net Juggler is an open source library that turns a commodity component cluster running the VR Juggler platform on each node into a single VR Juggler image cluster. Application parallelization is transparent to the user and leads to high performance executions even with limited bandwidth networks.

SoftGenLock: active stereo and genlock for PC cluster

In this paper, we present SoftGenLock, an open source software that enables genlock and active stereo on commodity graphics cards. SoftGenLock is implemented on top of Linux. It does not require any hardware modification of the graphics card. Rather than to gain total control on signal generation, which would make the software deeply dependent on the graphics card specification, SoftGenLock applies continuous small modifications to converge and maintain genlocked video signals. To be properly synchronized with each video retrace, SoftGenLock is executed as a real-time task. The genlock signal is propagated along the different machines using the parallel port, a low latency device present on all PCs. It results in a software that only requires access to few specific registers on a graphics card: it can be ported with minimal effort on potentially any graphics card.

Parallel Computing Flow Accumulation in Large Digital Elevation Models

This paper describes a new fast and scalable parallel algorithm to compute global flow accumulation for automatic drainage network extraction in large digital elevation models (DEM for short). Our method uses the D8 model to compute the flow directions for all pixels in the DEM (except NODATA and oceans). A parallel spanning tree algorithm is proposed to compute hierarchical catchment basins to model the flow of water from a sink (local minima) moving on DEM to its outlet (ocean, NODATA, or border of DEM). And finally, based on local flow accumulation and the hierarchical trees between sinks, we determinate entirely the global flow accumulation. From that, the drainage networks of DEM can be extracted. Our method does not need any preprocessing like stream burning on the initial DEM and tends to make the most of incomplete DEMs. Our algorithms are entirely parallel. Efficiency and scalability have been tested on different large DEMs.

A 3d data intensive tele-immersive grid

Networked virtual environments like Second Life enable distant people to meet for leisure as well as work. But users are represented through avatars controlled by keyboards and mouses, leading to a low sense of presence especially regarding body language. Multi-camera real-time 3D modeling offers a way to ensure a significantly higher sense of presence. But producing quality geometries, well textured, and to enable distant user tele-presence in non trivial virtual environments is still a challenge today. In this paper we present a tele-immersive system based on multi-camera 3D modeling. Users from distant sites are immersed in a rich virtual environment served by a parallel terrain rendering engine. Distant users, present through their 3D model, can perform some local interactions while having a strong visual presence. We experimented our system between three large cities a few hundreds kilometers apart from each other. This work demonstrate the feasibility of a rich 3D multimedia environment ensuring users a strong sense of presence.

Parallel computing of catchment basins in large digital elevation model

This paper presents a fast and flexible parallel implementation to compute catchment basins in the large digital elevation models (DEM for short). This algorithm aims at using all the specific properties of the problem to optimize local computations and to avoid useless communications or synchronizations. The algorithm has been implemented in MPI and the first benchmarks show the scalability of the method.

A Structured Synchronization and Communication Model Fitting Irregular Data Accesses

In this paper, we present a parallel programming and execution model based on alogicalordering of control flows. We show that it is possible to provide a unifying framework consisting of a synchronous programming model, thereby facilitating the mastery of programs, and an asynchronous execution model yielding efficient executions. Our approach is based on a SPMD and task parallel programming language, called SCL?Chan. Communications take place through channels and rely on explicit send/receive instructions. In contrast to classical message passing models, synchronizations and communications are dissociated. We show that it is possible to perform a data-driven automatic translation of sequential and arbitrary DOACROSS loops into SCL?Chan, by using nonmatching send/receive instructions. Our parallelization technique allows us to handle irregular control and leads to optimizations of communications in irregular computations.

Parallel computing of catchment basins of rivers in large digital elevation models

This paper describes a new fast and scalable parallel algorithm to automatically determine catchment basin of rivers in large digital elevation models (DEM for short). This algorithms is based on the construction of a minimal spanning tree, via a hierarchy of graphs, modeling the water route on the DEM. It does not need any preprocessing like stream burning on the initial DEM and tends to make the most of incomplete DEM. Efficiency and scalability have been tested on very large DEM.

Performance prediction for mappings of distributed applications on PC clusters

Distributed applications running on clusters may be composed of several components with very different performance requirements. The FlowVR middleware allows the developer to deploy such applications and to define communication and synchronization schemes between components without modifying the code. While it eases the creation of mappings, FlowVR does not come with a performance model. Consequently the optimization of mappings is left to the developers skills. But this task becomes difficult as the number of components and cluster nodes grow and even more complex if the cluster is composed of heterogeneous nodes and networks. In this paper we propose an approach to predict performance of FlowVR distributed applications given a mapping and a cluster. We also give some advice to the developer to create efficient mappings and to avoid configurations which may lead to unexpected performance. Since the FlowVR model is very close to underlying models of lots of distributed codes, our approach can be useful for all designers of such applications.

A scalable parallel minimum spanning tree algorithm for catchment basin delimitation in large digital elevation models

This paper describes a new fast and scalable parallel algorithm to automatically determine catchment basin of rivers in large digital elevation models (DEMs). This algorithm is based on the construction of a minimal spanning tree, via a hierarchy of graphs, modeling the water route on the DEM. It combines different techniques used in hydrogeology, image processing and graph theory to obtain the most accurate results in terms of geomorphology without any preprocessing. The method tends to exploit the most of the DEMs, avoiding misleading inconsistencies DEMs contain. It has been designed to be entirely parallel and scalable for architectures such as PC clusters. Some experiments are presented to show accuracy, efficiency and scalability on huge DEMs. Copyright