X-Git-Url: https://bilbo.iut-bm.univ-fcomte.fr/and/gitweb/rce2015.git/blobdiff_plain/d0129a1639a935ba009f36957b8ac7927c103a45..1bcc74bd1da4c5fcd944c3ee15de0e26ed3de258:/paper.tex?ds=inline diff --git a/paper.tex b/paper.tex index 34f7ec7..53d0dd9 100644 --- a/paper.tex +++ b/paper.tex @@ -21,7 +21,6 @@ \usepackage{algpseudocode} %\usepackage{amsthm} \usepackage{graphicx} -\usepackage[american]{babel} % Extension pour les liens intra-documents (tagged PDF) % et l'affichage correct des URL (commande \url{http://example.com}) %\usepackage{hyperref} @@ -75,23 +74,26 @@ analysis of simulated grid-enabled numerical iterative algorithms} %\itshape{\journalnamelc}\footnotemark[2]} -\author{ Charles Emile Ramamonjisoa and - David Laiymani and - Arnaud Giersch and - Lilia Ziane Khodja and - Raphaël Couturier +\author{Charles Emile Ramamonjisoa\affil{1}, + David Laiymani\affil{1}, + Arnaud Giersch\affil{1}, + Lilia Ziane Khodja\affil{2} and + Raphaël Couturier\affil{1} } \address{ - \centering - Femto-ST Institute - DISC Department\\ - Université de Franche-Comté\\ - Belfort\\ - Email: \email{{raphael.couturier,arnaud.giersch,david.laiymani,charles.ramamonjisoa}@univ-fcomte.fr} + \affilnum{1}% + Femto-ST Institute, DISC Department, + University of Franche-Comté, + Belfort, France. + Email:~\email{{charles.ramamonjisoa,david.laiymani,arnaud.giersch,raphael.couturier}@univ-fcomte.fr}\break + \affilnum{2} + Department of Aerospace \& Mechanical Engineering, + Non Linear Computational Mechanics, + University of Liege, Liege, Belgium. + Email:~\email{l.zianekhodja@ulg.ac.be} } -%% Lilia Ziane Khodja: Department of Aerospace \& Mechanical Engineering\\ Non Linear Computational Mechanics\\ University of Liege\\ Liege, Belgium. Email: l.zianekhodja@ulg.ac.be - \begin{abstract} The behavior of multi-core applications is always a challenge to predict, especially with a new architecture for which no experiment has been performed. With some applications, it is difficult, if not impossible, to build @@ -134,7 +136,7 @@ are often very important. So, in this context it is difficult to optimize a given application for a given architecture. In this way and in order to reduce the access cost to these computing resources it seems very interesting to use a simulation environment. The advantages are numerous: development life cycle, -code debugging, ability to obtain results quickly~\ldots. In counterpart, the simulation results need to be consistent with the real ones. +code debugging, ability to obtain results quickly\dots{} In counterpart, the simulation results need to be consistent with the real ones. In this paper we focus on a class of highly efficient parallel algorithms called \emph{iterative algorithms}. The parallel scheme of iterative methods is quite @@ -163,10 +165,11 @@ application on a given multi-core architecture. Finding good resource allocations policies under varying CPU power, network speeds and loads is very challenging and labor intensive~\cite{Calheiros:2011:CTM:1951445.1951450}. This problematic is even more difficult for the asynchronous scheme where a small -parameter variation of the execution platform can lead to very different numbers -of iterations to reach the converge and so to very different execution times. In -this challenging context we think that the use of a simulation tool can greatly -leverage the possibility of testing various platform scenarios. +parameter variation of the execution platform and of the application data can +lead to very different numbers of iterations to reach the converge and so to +very different execution times. In this challenging context we think that the +use of a simulation tool can greatly leverage the possibility of testing various +platform scenarios. The main contribution of this paper is to show that the use of a simulation tool (i.e. the SimGrid toolkit~\cite{SimGrid}) in the context of real parallel @@ -174,10 +177,33 @@ applications (i.e. large linear system solvers) can help developers to better tune their application for a given multi-core architecture. To show the validity of this approach we first compare the simulated execution of the multisplitting algorithm with the GMRES (Generalized Minimal Residual) -solver~\cite{saad86} in synchronous mode. The obtained results on different -simulated multi-core architectures confirm the real results previously obtained -on non simulated architectures. We also confirm the efficiency of the -asynchronous multisplitting algorithm compared to the synchronous GMRES. In +solver~\cite{saad86} in synchronous mode. The simulation results allow us to +determine which method to choose given a specified multi-core architecture. + +\LZK{Pas trop convainquant comme argument pour valider l'approche de simulation. \\On peut dire par exemple: on a pu simuler différents algos itératifs à large échelle (le plus connu GMRES et deux variantes de multisplitting) et la simulation nous a permis (sans avoir le vrai matériel) de déterminer quelle serait la meilleure solution pour une telle configuration de l'archi ou vice versa.\\A revoir...} +\DL{OK : ajout d'une phrase précisant tout cela} + +Moreover the obtained results on different simulated multi-core architectures +confirm the real results previously obtained on non simulated architectures. +More precisely the simulated results are in accordance (i.e. with the same order +of magnitude) with the works presented in [], which show that the synchronous +multisplitting method is more efficient than GMRES for large scale clusters. + +\LZK{Il n y a pas dans la partie expé cette comparaison et confirmation des +résultats entre la simulation et l'exécution réelle des algos sur les vrais +clusters.\\ Sinon on pourrait ajouter dans la partie expé une référence vers le +journal supercomput de krylov multi pour confirmer que cette méthode est +meilleure que GMRES sur les clusters large échelle.} \DL{OK ajout d'une phrase. +Par contre je n'ai pas la ref. Merci de la mettre} + +Simulated results also confirm the efficiency of the asynchronous +multisplitting algorithm compared to the synchronous GMRES especially in case of +geographically distant clusters. + +\LZK{P.S.: Pour tout le papier, le principal objectif n'est pas de faire des comparaisons entre des méthodes itératives!!\\Sinon, les deux algorithmes Krylov multisplitting synchrone et multisplitting asynchrone sont plus efficaces que GMRES sur des clusters à large échelle.\\Et préciser, si c'est vraiment le cas, que le multisplitting asynchrone est plus efficace et adapté aux clusters distants par rapport aux deux autres algos (je n'ai pas encore lu la partie expé)} +\DL{Tu as raison on s'est posé la question de garder ou non cette partie des résultats. On a décidé de la garder pour avoir plus de chose à montrer. J'ai essayer de clarifier un peu} + +In this way and with a simple computing architecture (a laptop) SimGrid allows us to run a test campaign of a real parallel iterative applications on different simulated multi-core architectures. To our knowledge, there is no @@ -191,8 +217,10 @@ Section~\ref{sec:04} details the different solvers that we use. Finally our experimental results are presented in section~\ref{sec:expe} followed by some concluding remarks and perspectives. +\LZK{Proposition d'un titre pour le papier: Grid-enabled simulation of large-scale linear iterative solvers.} + -\section{The asynchronous iteration model} +\section{The asynchronous iteration model and the motivations of our work} \label{sec:asynchro} Asynchronous iterative methods have been studied for many years theoritecally and @@ -216,10 +244,77 @@ point. In the asynchronous model, the convergence detection is more tricky as it must not synchronize all the processors. Interested readers can consult~\cite{myBCCV05c,bahi07,ccl09:ij}. +The number of iterations required to reach the convergence is generally greater +for the asynchronous scheme (this number depends depends on the delay of the +messages). Note that, it is not the case in the synchronous mode where the +number of iterations is the same than in the sequential mode. In this way, the +set of the parameters of the platform (number of nodes, power of nodes, +inter and intra clusters bandwidth and latency, \ldots) and of the +application can drastically change the number of iterations required to get the +convergence. It follows that asynchronous iterative algorithms are difficult to +optimize since the financial and deployment costs on large scale multi-core +architecture are often very important. So, prior to delpoyment and tests it +seems very promising to be able to simulate the behavior of asynchronous +iterative algorithms. The problematic is then to show that the results produce +by simulation are in accordance with reality i.e. of the same order of +magnitude. To our knowledge, there is no study on this problematic. + \section{SimGrid} - \label{sec:simgrid} +\label{sec:simgrid} +SimGrid~\cite{SimGrid,casanova+legrand+quinson.2008.simgrid,casanova+giersch+legrand+al.2014.versatile} is a discrete event simulation framework to study the behavior of large-scale distributed computing platforms as Grids, Peer-to-Peer systems, Clouds and High Performance Computation systems. It is widely used to simulate and evaluate heuristics, prototype applications or even assess legacy MPI applications. It is still actively developed by the scientific community and distributed as an open source software. %%%%%%%%%%%%%%%%%%%%%%%%% +% SimGrid~\cite{SimGrid,casanova+legrand+quinson.2008.simgrid,casanova+giersch+legrand+al.2014.versatile} +% is a simulation framework to study the behavior of large-scale distributed +% systems. As its name suggests, it emanates from the grid computing community, +% but is nowadays used to study grids, clouds, HPC or peer-to-peer systems. The +% early versions of SimGrid date back from 1999, but it is still actively +% developed and distributed as an open source software. Today, it is one of the +% major generic tools in the field of simulation for large-scale distributed +% systems. + +SimGrid provides several programming interfaces: MSG to simulate Concurrent +Sequential Processes, SimDAG to simulate DAGs of (parallel) tasks, and SMPI to +run real applications written in MPI~\cite{MPI}. Apart from the native C +interface, SimGrid provides bindings for the C++, Java, Lua and Ruby programming +languages. SMPI is the interface that has been used for the work described in +this paper. The SMPI interface implements about \np[\%]{80} of the MPI 2.0 +standard~\cite{bedaride+degomme+genaud+al.2013.toward}, and supports +applications written in C or Fortran, with little or no modifications (cf Section IV - paragraph B). + +Within SimGrid, the execution of a distributed application is simulated by a +single process. The application code is really executed, but some operations, +like communications, are intercepted, and their running time is computed +according to the characteristics of the simulated execution platform. The +description of this target platform is given as an input for the execution, by +means of an XML file. It describes the properties of the platform, such as +the computing nodes with their computing power, the interconnection links with +their bandwidth and latency, and the routing strategy. The scheduling of the +simulated processes, as well as the simulated running time of the application +are computed according to these properties. + +To compute the durations of the operations in the simulated world, and to take +into account resource sharing (e.g. bandwidth sharing between competing +communications), SimGrid uses a fluid model. This allows users to run relatively fast +simulations, while still keeping accurate +results~\cite{bedaride+degomme+genaud+al.2013.toward, + velho+schnorr+casanova+al.2013.validity}. Moreover, depending on the +simulated application, SimGrid/SMPI allows to skip long lasting computations and +to only take their duration into account. When the real computations cannot be +skipped, but the results are unimportant for the simulation results, it is +also possible to share dynamically allocated data structures between +several simulated processes, and thus to reduce the whole memory consumption. +These two techniques can help to run simulations on a very large scale. + +The validity of simulations with SimGrid has been asserted by several studies. +See, for example, \cite{velho+schnorr+casanova+al.2013.validity} and articles +referenced therein for the validity of the network models. Comparisons between +real execution of MPI applications on the one hand, and their simulation with +SMPI on the other hand, are presented in~\cite{guermouche+renard.2010.first, + clauss+stillwell+genaud+al.2011.single, + bedaride+degomme+genaud+al.2013.toward}. All these works conclude that +SimGrid is able to simulate pretty accurately the real behavior of the +applications. %%%%%%%%%%%%%%%%%%%%%%%%% \section{Two-stage multisplitting methods} @@ -241,7 +336,7 @@ where $x_\ell$ are sub-vectors of the solution $x$, $b_\ell$ are the sub-vectors A_{\ell\ell} x_\ell = c_\ell,\mbox{~for~}\ell=1,\ldots,L, \label{eq:03} \end{equation} -where right-hand sides $c_\ell=b_\ell-\sum_{m\neq\ell}A_{\ell m}x_m$ are computed using the shared vectors $x_m$. In this paper, we use the well-known iterative method GMRES ({\it Generalized Minimal RESidual})~\cite{saad86} as an inner iteration to approximate the solutions of the different splittings arising from the block Jacobi multisplitting of matrix $A$. The algorithm in Figure~\ref{01} shows the main key points of our block Jacobi two-stage method executed by a cluster of processors. In line~\ref{solve}, the linear sub-system~(\ref{eq:03}) is solved in parallel using GMRES method where $\MIG$ and $\TOLG$ are the maximum number of inner iterations and the tolerance threshold for GMRES respectively. The convergence of the two-stage multisplitting methods, based on synchronous or asynchronous iterations, has been studied by many authors for example~\cite{Bru95,bahi07}. +where right-hand sides $c_\ell=b_\ell-\sum_{m\neq\ell}A_{\ell m}x_m$ are computed using the shared vectors $x_m$. In this paper, we use the well-known iterative method GMRES ({\it Generalized Minimal RESidual})~\cite{saad86} as an inner iteration to approximate the solutions of the different splittings arising from the block Jacobi multisplitting of matrix $A$. The algorithm in Figure~\ref{alg:01} shows the main key points of our block Jacobi two-stage method executed by a cluster of processors. In line~\ref{solve}, the linear sub-system~(\ref{eq:03}) is solved in parallel using GMRES method where $\MIG$ and $\TOLG$ are the maximum number of inner iterations and the tolerance threshold for GMRES respectively. The convergence of the two-stage multisplitting methods, based on synchronous or asynchronous iterations, has been studied by many authors for example~\cite{Bru95,bahi07}. \begin{figure}[t] %\begin{algorithm}[t] @@ -357,8 +452,6 @@ In addition, the following arguments are given to the programs at runtime: \item maximum number of restarts for the Arnorldi process in GMRES method, \item execution mode: synchronous or asynchronous. \end{itemize} -\LZK{CE pourrais tu vérifier et confirmer les valeurs des éléments diag et off-diag de la matrice?} -\RCE{oui, les valeurs de diag et off-diag donnees sont ok} It should also be noticed that both solvers have been executed with the Simgrid selector \texttt{-cfg=smpi/running\_power} which determines the computational power (here 19GFlops) of the simulator host machine. @@ -467,7 +560,7 @@ and between distant clusters. This parameter is application dependent. In the scope of this paper, our first objective is to analyze when the Krylov Multisplitting method has better performance than the classical GMRES -method. With a synchronous iterative method, better performance mean a +method. With a synchronous iterative method, better performance means a smaller number of iterations and execution time before reaching the convergence. For a systematic study, the experiments should figure out that, for various grid parameters values, the simulator will confirm the targeted outcomes, @@ -492,7 +585,8 @@ architectures and scaling up the input matrix size} Input matrix size & N$_{x}$ x N$_{y}$ x N$_{z}$ =150 x 150 x 150\\ %\hline - & N$_{x}$ x N$_{y}$ x N$_{z}$ =170 x 170 x 170 \\ \hline \end{tabular} -\caption{Test conditions: Various grid configurations with the input matix size N$_{x}$=150 or N$_{x}$=170 \RC{je ne comprends pas la légende... Ca ne serait pas plutot Characteristics of cluster (mais il faudrait lui donner un nom)}} +\caption{Test conditions: various grid configurations with the input matix size N$_{x}$=150 or N$_{x}$=170 \RC{N2 n'est pas défini..}\RC{Nx est défini, Ny? Nz?} +\AG{La lettre 'x' n'est pas le symbole de la multiplication. Utiliser \texttt{\textbackslash times}. Idem dans le texte, les figures, etc.}} \label{tab:01} \end{center} \end{table} @@ -500,8 +594,6 @@ architectures and scaling up the input matrix size} -%\RCE{J'ai voulu mettre les tableaux des données mais je pense que c'est inutile et ça va surcharger} - In this section, we analyze the performance of algorithms running on various grid configurations (2x16, 4x8, 4x16 and 8x8). First, the results in Figure~\ref{fig:01} @@ -517,7 +609,8 @@ multisplitting method. \begin{center} \includegraphics[width=100mm]{cluster_x_nodes_nx_150_and_nx_170.pdf} \end{center} - \caption{Various grid configurations with the input matrix size N$_{x}$=150 and N$_{x}$=170} + \caption{Various grid configurations with the input matrix size N$_{x}$=150 and N$_{x}$=170\RC{idem} +\AG{Utiliser le point comme séparateur décimal et non la virgule. Idem dans les autres figures.}} \label{fig:01} \end{figure} @@ -527,9 +620,9 @@ grid architectures, even with the same number of processors (for example, 2x16 and 4x8). We can observ the low sensitivity of the Krylov multisplitting method (compared with the classical GMRES) when scaling up the number of the processors in the grid: in average, the GMRES (resp. Multisplitting) algorithm performs -$40\%$ better (resp. $48\%$) when running from 2x16=32 to 8x8=64 processors. +$40\%$ better (resp. $48\%$) when running from 2x16=32 to 8x8=64 processors. \RC{pas très clair, c'est pas précis de dire qu'un algo perform mieux qu'un autre, selon quel critère?} -\subsubsection{Running on two different inter-clusters network speeds \\} +\subsubsection{Running on two different inter-clusters network speeds \\} \begin{table} [ht!] \begin{center} @@ -540,17 +633,16 @@ $40\%$ better (resp. $48\%$) when running from 2x16=32 to 8x8=64 processors. - & N2 : bw=1Gbs-lat=5.10$^{-5}$ \\ Input matrix size & N$_{x}$ x N$_{y}$ x N$_{z}$ =150 x 150 x 150\\ \hline \end{tabular} -\caption{Test conditions: Grid 2x16 and 4x8 - Networks N1 vs N2} +\caption{Test conditions: grid 2x16 and 4x8 with networks N1 vs N2} \label{tab:02} \end{center} \end{table} These experiments compare the behavior of the algorithms running first on a -speed inter-cluster network (N1) and also on a less performant network (N2). -Figure~\ref{fig:02} shows that end users will gain to reduce the execution time -for both algorithms in using a grid architecture like 4x16 or 8x8: the -performance was increased by a factor of $2$. The results depict also that when -the network speed drops down (variation of 12.5\%), the difference between the two Multisplitting algorithms execution times can reach more than 25\%. +speed inter-cluster network (N1) and also on a less performant network (N2). \RC{Il faut définir cela avant...} +Figure~\ref{fig:02} shows that end users will reduce the execution time +for both algorithms when using a grid architecture like 4x16 or 8x8: the reduction is about $2$. The results depict also that when +the network speed drops down (variation of 12.5\%), the difference between the two Multisplitting algorithms execution times can reach more than 25\%. %\RC{c'est pas clair : la différence entre quoi et quoi?} %\DL{pas clair} %\RCE{Modifie} @@ -560,7 +652,8 @@ the network speed drops down (variation of 12.5\%), the difference between t \begin{figure} [ht!] \centering \includegraphics[width=100mm]{cluster_x_nodes_n1_x_n2.pdf} -\caption{Grid 2x16 and 4x8 - Networks N1 vs N2} +\caption{Grid 2x16 and 4x8 with networks N1 vs N2 +\AG{\np{8E-6}, \np{5E-6} au lieu de 8E-6, 5E-6}} \label{fig:02} \end{figure} %\end{wrapfigure} @@ -576,7 +669,7 @@ the network speed drops down (variation of 12.5\%), the difference between t Network & N1 : bw=1Gbs \\ %\hline Input matrix size & N$_{x}$ x N$_{y}$ x N$_{z}$ =150 x 150 x 150\\ \hline \end{tabular} -\caption{Test conditions: Network latency impacts} +\caption{Test conditions: network latency impacts} \label{tab:03} \end{table} @@ -585,20 +678,22 @@ the network speed drops down (variation of 12.5\%), the difference between t \begin{figure} [ht!] \centering \includegraphics[width=100mm]{network_latency_impact_on_execution_time.pdf} -\caption{Network latency impacts on execution time} +\caption{Network latency impacts on execution time +\AG{\np{E-6}}} \label{fig:03} \end{figure} -According to the results of Figure~\ref{fig:03}, a degradation of the network -latency from $8.10^{-6}$ to $6.10^{-5}$ implies an absolute time increase of more -than $75\%$ (resp. $82\%$) of the execution for the classical GMRES (resp. Krylov -multisplitting) algorithm. In addition, it appears that the Krylov -multisplitting method tolerates more the network latency variation with a less -rate increase of the execution time. Consequently, in the worst case -($lat=6.10^{-5 }$), the execution time for GMRES is almost the double than the -time of the Krylov multisplitting, even though, the performance was on the same -order of magnitude with a latency of $8.10^{-6}$. +According to the results of Figure~\ref{fig:03}, a degradation of the network +latency from $8.10^{-6}$ to $6.10^{-5}$ implies an absolute time increase of +more than $75\%$ (resp. $82\%$) of the execution for the classical GMRES +(resp. Krylov multisplitting) algorithm. In addition, it appears that the +Krylov multisplitting method tolerates more the network latency variation with a +less rate increase of the execution time.\RC{Les 2 précédentes phrases me + semblent en contradiction....} Consequently, in the worst case ($lat=6.10^{-5 +}$), the execution time for GMRES is almost the double than the time of the +Krylov multisplitting, even though, the performance was on the same order of +magnitude with a latency of $8.10^{-6}$. \subsubsection{Network bandwidth impacts on performance} \ \\ @@ -610,7 +705,7 @@ order of magnitude with a latency of $8.10^{-6}$. Network & N1 : bw=1Gbs - lat=5.10$^{-5}$ \\ %\hline Input matrix size & N$_{x}$ x N$_{y}$ x N$_{z}$ =150 x 150 x 150\\ \hline \\ \end{tabular} -\caption{Test conditions: Network bandwidth impacts} +\caption{Test conditions: Network bandwidth impacts\RC{Qu'est ce qui varie ici? Il n'y a pas de variation dans le tableau}} \label{tab:04} \end{table} @@ -618,7 +713,8 @@ order of magnitude with a latency of $8.10^{-6}$. \begin{figure} [ht!] \centering \includegraphics[width=100mm]{network_bandwith_impact_on_execution_time.pdf} -\caption{Network bandwith impacts on execution time} +\caption{Network bandwith impacts on execution time +\AG{``Execution time'' avec un 't' minuscule}. Idem autres figures.} \label{fig:04} \end{figure} @@ -656,9 +752,9 @@ In these experiments, the input matrix size has been set from $N_{x} = N_{y} time for both algorithms increases when the input matrix size also increases. But the interesting results are: \begin{enumerate} - \item the drastic increase ($10$ times) \RC{Je ne vois pas cela sur la figure} -\RCE{Corrige} of the number of iterations needed to reach the convergence for the classical -GMRES algorithm when the matrix size go beyond $N_{x}=150$; + \item the drastic increase ($10$ times) of the number of iterations needed to + reach the convergence for the classical GMRES algorithm when the matrix size + go beyond $N_{x}=150$; \RC{C'est toujours pas clair... ok le nommbre d'itérations est 10 fois plus long mais la suite de la phrase ne veut rien dire} \item the classical GMRES execution time is almost the double for $N_{x}=140$ compared with the Krylov multisplitting method. \end{enumerate} @@ -786,7 +882,8 @@ geographically distant clusters through the internet. \hline \end{mytable} %\end{table} - \caption{Relative gain of the multisplitting algorithm compared with the classical GMRES} + \caption{Relative gain of the multisplitting algorithm compared with the classical GMRES +\AG{C'est un tableau, pas une figure}} \label{fig:07} \end{figure} @@ -795,13 +892,15 @@ geographically distant clusters through the internet. CONCLUSION -\section*{Acknowledgment} - +%\section*{Acknowledgment} +\ack This work is partially funded by the Labex ACTION program (contract ANR-11-LABX-01-01). - \bibliographystyle{wileyj} \bibliography{biblio} +\AG{Warning bibtex à corriger (% + \texttt{empty booktitle in Bru95}% +).} \end{document}