HPCC2014

[hpcc2014.git] / hpcc.tex

diff --git a/hpcc.tex b/hpcc.tex
index 4ff46b6d368fe5a7b4efe4b8f0f994baf496a3c6..cbd1e9ca4f74e58b240b1ea250c9a756d69671c7 100644 (file)
--- a/hpcc.tex
+++ b/hpcc.tex
@@ -10,7 +10,7 @@
 \usepackage{graphicx}
 \usepackage[american]{babel}
 % Extension pour les liens intra-documents (tagged PDF)
 \usepackage{graphicx}
 \usepackage[american]{babel}
 % Extension pour les liens intra-documents (tagged PDF)

-% et l'affichage correct des URL (commande \url{http://example.com})
+% et l'affichage correct des UR (commande \url{http://example.com})

 %\usepackage{hyperref}
 
 \usepackage{url}
 %\usepackage{hyperref}
 
 \usepackage{url}


@@ -182,7 +182,7 @@ in many scientific domains.  They can be classified in three main classes
 depending on how iterations and communications are managed (for more details
 readers can refer to~\cite{bcvc06:ij}). In the synchronous iterations model,
 data are exchanged at the end of each iteration. All the processors must begin
 depending on how iterations and communications are managed (for more details
 readers can refer to~\cite{bcvc06:ij}). In the synchronous iterations model,
 data are exchanged at the end of each iteration. All the processors must begin

-the same iteration at the same time and important idle times on processors are
+the same iteration at the same time and important and useless idle times used for synchronization on processors are

 generated.  It is possible to use asynchronous communications, in this case, the
 model can be compared to the previous one except that data required on another
 processor are sent asynchronously i.e.  without stopping current computations.
 generated.  It is possible to use asynchronous communications, in this case, the
 model can be compared to the previous one except that data required on another
 processor are sent asynchronously i.e.  without stopping current computations.


@@ -261,7 +261,7 @@ 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
 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.
+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,
 
 Within SimGrid, the execution of a distributed application is simulated by a
 single process.  The application code is really executed, but some operations,


@@ -416,11 +416,11 @@ cluster 1 broadcasts a stop message to the masters of other clusters. In this wo
 the local convergence on each cluster $\ell$ is detected when the following
 condition is satisfied
 \begin{equation*}
 the local convergence on each cluster $\ell$ is detected when the following
 condition is satisfied
 \begin{equation*}

-  (k\leq \MI) \text{ or } (\|X_\ell^k - X_\ell^{k+1}\|_{\infty}\leq\epsilon)
+  (k=\MI) \text{ or } (\|X_\ell^k - X_\ell^{k+1}\|_{\infty}\leq\epsilon)

 \end{equation*}
 where $\MI$ is the maximum number of outer iterations and $\epsilon$ is the
 tolerance threshold of the error computed between two successive local solution
 \end{equation*}
 where $\MI$ is the maximum number of outer iterations and $\epsilon$ is the
 tolerance threshold of the error computed between two successive local solution

-$X_\ell^k$ and $X_\ell^{k+1}$.
+$X_\ell^k$ and $X_\ell^{k+1}$. It should be noted that with asynchronous iterative algorithms, we cannot use a classical norm (which would require to synchronize all processors), such as $\|X_\ell^k - X_\ell^{k+1}\|_{2}$ for example. Nevertheless, in our experiments, we check that the final result is correct, for this we compute the precision with $max_i | A*x-b |_i$.

 
 
 
 
 
 


@@ -461,8 +461,8 @@ The parallel solving of the 3D Poisson problem with our multisplitting method re
 
 
 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
 
 
 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

-We did not encounter major blocking problems when adapting the multisplitting algorithm previously described to a simulation environment like SimGrid unless some code 
-debugging. Indeed, apart from the review of the program sequence for asynchronous exchanges between processors within a cluster or between clusters, the algorithm was executed successfully with SMPI and provided identical outputs as those obtained with direct execution under MPI. For the synchronous GMRES method, the execution of the program raised no particular issue but in the asynchronous multisplitting method, the review of the sequence of \texttt{MPI\_Isend, MPI\_Irecv} and \texttt{MPI\_Waitall} instructions
+We did not encounter major blocking problems when adapting the multisplitting algorithm previously described to a simulation environment like SimGrid. Only, some code 
+debugging has been required. Indeed, apart from the review of the program sequence for asynchronous exchanges between processors within a cluster or between clusters, the algorithm was executed successfully with SMPI and provided identical outputs as those obtained with direct execution under MPI. For the synchronous GMRES method, the execution of the program raised no particular issue but in the asynchronous multisplitting method, the review of the sequence of \texttt{MPI\_Isend, MPI\_Irecv} and \texttt{MPI\_Waitall} instructions

 and with the addition of the primitive \texttt{MPI\_Test} was needed to avoid a memory fault due to an infinite loop resulting from the non-convergence of the algorithm.
 %\CER{On voulait en fait montrer la simplicité de l'adaptation de l'algo a SimGrid. Les problèmes rencontrés décrits dans ce paragraphe concerne surtout le mode async}\LZK{OK. J'aurais préféré avoir un peu plus de détails sur l'adaptation de la version async} 
 %\CER{Le problème majeur sur l'adaptation MPI vers SMPI pour la partie asynchrone de l'algorithme a été le plantage en SMPI de Waitall après un Isend et Irecv. J'avais proposé un workaround en utilisant un MPI\_wait séparé pour chaque échange a la place d'un waitall unique pour TOUTES les échanges, une instruction qui semble bien fonctionner en MPI. Ce workaround aussi fonctionne bien. Mais après, tu as modifié le programme avec l'ajout d'un MPI\_Test, au niveau de la routine de détection de la convergence et du coup, l'échange global avec waitall a aussi fonctionné.}
 and with the addition of the primitive \texttt{MPI\_Test} was needed to avoid a memory fault due to an infinite loop resulting from the non-convergence of the algorithm.
 %\CER{On voulait en fait montrer la simplicité de l'adaptation de l'algo a SimGrid. Les problèmes rencontrés décrits dans ce paragraphe concerne surtout le mode async}\LZK{OK. J'aurais préféré avoir un peu plus de détails sur l'adaptation de la version async} 
 %\CER{Le problème majeur sur l'adaptation MPI vers SMPI pour la partie asynchrone de l'algorithme a été le plantage en SMPI de Waitall après un Isend et Irecv. J'avais proposé un workaround en utilisant un MPI\_wait séparé pour chaque échange a la place d'un waitall unique pour TOUTES les échanges, une instruction qui semble bien fonctionner en MPI. Ce workaround aussi fonctionne bien. Mais après, tu as modifié le programme avec l'ajout d'un MPI\_Test, au niveau de la routine de détection de la convergence et du coup, l'échange global avec waitall a aussi fonctionné.}


@@ -683,7 +683,7 @@ stable even if the residual error precision varies from \np{E-5} to \np{E-9}. By
 increasing the matrix size up to $100^3$ elements, it was necessary to increase the
 CPU power by \np[\%]{50} to \np[GFlops]{1.5} to get the algorithm convergence and the same order of asynchronous mode efficiency.  Maintaining  a relative gain of $2.5$ and such processor power but increasing network throughput inter cluster up to \np[Mbit/s]{50},  is obtained with
 high external precision of \np{E-11} for a matrix size from $110^3$ to $150^3$ side
 increasing the matrix size up to $100^3$ elements, it was necessary to increase the
 CPU power by \np[\%]{50} to \np[GFlops]{1.5} to get the algorithm convergence and the same order of asynchronous mode efficiency.  Maintaining  a relative gain of $2.5$ and such processor power but increasing network throughput inter cluster up to \np[Mbit/s]{50},  is obtained with
 high external precision of \np{E-11} for a matrix size from $110^3$ to $150^3$ side

-elements.
+elements. 

 
 %For the 3 clusters architecture including a total of 100 hosts,
 %Table~\ref{tab.cluster.3x33} shows that it was difficult to have a combination
 
 %For the 3 clusters architecture including a total of 100 hosts,
 %Table~\ref{tab.cluster.3x33} shows that it was difficult to have a combination


@@ -706,7 +706,7 @@ elements.
 %\CER{Définitivement, les paramètres réseaux variables ici se rapportent au réseau INTER cluster.}
 \section{Conclusion}
 The simulation of the execution of parallel asynchronous iterative algorithms on large scale  clusters has been presented. 
 %\CER{Définitivement, les paramètres réseaux variables ici se rapportent au réseau INTER cluster.}
 \section{Conclusion}
 The simulation of the execution of parallel asynchronous iterative algorithms on large scale  clusters has been presented. 

-In this work, we show that SimGrid is an efficient simulation tool that has enabled us to 
+In this work, we show that SimGrid is one of efficient simulation tool that has enabled us to 

 reach the following two objectives: 
 
 \begin{enumerate}
 reach the following two objectives: 
 
 \begin{enumerate}