Distributed simulation of parallel computers

نویسندگان

  • Loïc Prylli
  • Bernard Tourancheau
چکیده

We propose a method for emulating the behavior of your favorite MIMD computer on a network of workstations, so that programs written for the MIMD target can run on the distributed emulated version with just a recompilation. This is achieved by executing the computational parts as in the native run, with the message passing being simulated. The hardware of the target machine is simulated so that the behavior of the application is identical to a native run on the simulated computer with virtual timings and trace les. Moreover, our complexity analysis sets up the conditions required to achieve a good speedup as a function of the number of simulation hosts, the network latency and the granularity of the application. Simulation distribu ee d'ordinateurs parall eles R esum e : Cet article pr esente nos travaux sur la simulation distribu ee, dirig ee par les ev enements, d'ordinateurs parall eles a m emoire distribu ee. Nous d ecrivons nos algorithmes de simulation distribu ee et l'analyse th eorique des conditions n ecessaires pour obtenir une bonne acc el eration de la simulation. Notre impl ementation d'un tel simulateur permet, a partir d'une application ecrite pour un ordinateur MIMD, une ex ecution sur un ensemble de stations de travail avec seulement une recompilation du code. La machine cible est simul ee au niveau mat eriel de mani ere a ce que le comportement de l'application soit le m^ eme que sur l'ordinateur simul e avec le maintient de temps d'ex ecution virtuels et la construction des chiers de traces correspondants. Les r esultats obtenus corroborent notre etude de complexit e et sugg erent le nombre de processeurs de la machine simulante, la granularit e de l'application pa-rall ele qui permettront d'obtenir une bonne eecacit e de la simulation parall ele.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Parallel Spatial Pyramid Match Kernel Algorithm for Object Recognition using a Cluster of Computers

This paper parallelizes the spatial pyramid match kernel (SPK) implementation. SPK is one of the most usable kernel methods, along with support vector machine classifier, with high accuracy in object recognition. MATLAB parallel computing toolbox has been used to parallelize SPK. In this implementation, MATLAB Message Passing Interface (MPI) functions and features included in the toolbox help u...

متن کامل

Turbomachinery CFD on Parallel Computers

The role of multistage turbomachinery simulation in the development of propulsion system models is discussed. Particularly, the need for simulations with higher fidelity and faster turnaround time is highlighted. It is shown how such fast simulations can be used in engineering-oriented environments. The use of parallel processing to achieve the required turnaround times is discussed. Current wo...

متن کامل

A Technique for the Distributed Simulation of Parallel Computers

A new technique for the efficient asynchronous discrete eventdriven simulation of parallel shared-memory computers is proposed. Our execution-driven methodology, while introducing minimal synchronization overhead to maintain a coherent distributed event causality relation, allows complete virtualization of the design at all levels and therefore is very flexible. We give a detailed description o...

متن کامل

Performance analysis of direct N-body algorithms for astrophysical simulations on distributed systems

We discuss the performance of direct summation codes used in the simulation of astrophysical stellar systems on highly distributed architectures. These codes compute the gravitational interaction among stars in an exact way and have an O(N2) scaling with the number of particles. They can be applied to a variety of astrophysical problems, like the evolution of star clusters, the dynamics of blac...

متن کامل

High-level Simulation of BSP Architectures

A major challenge for parallel computing to become a widespread discipline is the development of a standardized combination of portable and efficient parallel programming. A major trend within computer architecture is that the most powerful computers are distributed memory machines with message passing. The explicit use of message passing in parallel programs may be considered as the assembler ...

متن کامل

Distributed BEAGLE: An Environment for Parallel and Distributed Evolutionary Computations

Evolutionary computation is a promising artificial intelligence field involving the simulation of natural evolution to solve problems. Given its implicit parallelism and high computational requirements, evolutionary computation is the perfect candidate for high performance parallel computers. This paper presents Distributed BEAGLE, a new master-slave architecture for parallel and distributed ev...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 1996