Merge branch 'main' into 77-fix-ecole-polytechnique-logo

chapters/software.tex

@@ -1,6 +1,6 @@
 \clearpage
 \chapter{Software}
-\label{sec:software}
+\label{chap:software}
 
 This chapter presents the software developed within Exa-MA, focusing on features, mathematics, functionalities, publications, acknowledgments, and contact details.
 

exa-ma-d7.1.tex

@@ -205,6 +205,7 @@
 %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
 
 \chapter{Introduction}
+\label{chap:introduction}
 \input{sections/summary}
 \input{sections/introduction}
 \import{chapters/}{benchmarking}

sections/introduction.tex

@@ -43,7 +43,7 @@ \subsection{Structure of the Document}
 
 \begin{itemize}
     \item \Cref{chap:introduction} introduces the purpose and scope of the deliverable, providing an overview of the Exa-MA project.
-    \item \Cref{chap:benchmarking} presents the benchmarking methodology that will be developed for exascale software, detailing the key objectives, testing processes, data management strategies, fault tolerance mechanisms, and the use of demonstrators.
+    \item \Cref{chap:methodology} presents the benchmarking methodology that will be developed for exascale software, detailing the key objectives, testing processes, data management strategies, fault tolerance mechanisms, and the use of demonstrators.
     \item \Cref{chap:software} provides a general overview of the software developed within Exa-MA, focusing on their features, mathematical foundations, functionalities, relevant publications, acknowledgments, and contact details. It includes general statistics about the software, offering insights into their characteristics and technological choices, such as supported architectures, programming languages, parallelism technologies, data formats, and DevOps practices.
     \item \Cref{chap:wp1} (\textbf{WP1 - Discretization}) presents software relevant to the discretization methods developed within this work package. For each software, we detail their features, the methods they implement, parallel capabilities, benchmarks and metrics developed, challenges identified, and provide a 12-month roadmap.
     \item \Cref{chap:wp2} (\textbf{WP2 - Model Order, Surrogate, Scientific Machine Learning Methods}) covers software related to reduced-order models, surrogate modeling, and scientific machine learning techniques tailored for exascale applications. Each software is presented with its implemented methods, parallel capabilities, associated benchmarks and metrics, challenges identified, and a 12-month roadmap.

software/hawen/WP1/WP1.tex

@@ -234,10 +234,15 @@ \subsection{12-Month Roadmap}
 
 
 
-We wish to improve input/output data management by including more options 
+We wish to standardize the use of high-order polynomials, which 
+is critical for the HDG method. By high-order, we mean at least 
+order 7 in three dimensions, so that the static condensation used
+by the method becomes more useful. For instance, using high orders 
+to have robust and efficient quadrature rules.
+In addition, we wish to improve input/output data management by including more options 
 to save the results of \hawen, for instance this can include the hdf5 format
 which is not allowed currently.
-In addition, the configuration of the benchmark and the results obtained will 
+Eventually, the configuration of the benchmark and the results obtained will 
 be made available in open-access for reproducibility.
 %In this section, describe the roadmap for improving benchmarks and addressing the challenges identified. This should include:
 %\begin{itemize}
@@ -263,10 +268,10 @@ \subsection{12-Month Roadmap}
 
     \rowcolor{numpexgray}{\rule{0pt}{2.5ex}\color{white}\bf Bottlenecks} &  {\rule{0pt}{2.5ex}\color{white}\bf Short Description }\\ 
 
-\rowcolor{white}    B10 - Scientific Productivity & provide short description here \\
-\rowcolor{numpexlightergray}    B11 - Reproducibility and Replicability of Computation & provide short description here \\
-\rowcolor{white}    B6 - Data Management & provide short description here \\
-\rowcolor{numpexlightergray}    B7 - Exascale Algorithms & provide short description here \\
+\rowcolor{white}    B10 - Scientific Productivity & Improve high-order polynomials operations for HDG discretization. \\
+\rowcolor{numpexlightergray}    B11 - Reproducibility and Replicability of Computation & Each benchmark will have a separate repository with the input files and associated results to reproduce the simulations. \\
+\rowcolor{white}    B6 - Data Management & Provide the meshes used for the benchmark and the associated results in standard format such as vtk and hdf5. \\
+\rowcolor{numpexlightergray}    B7 - Exascale Algorithms & Control the memory footprint. \\
 \end{tabular}
         }
     }

software/hawen/WP3/WP3.tex

@@ -188,10 +188,17 @@ \subsection{12-Month Roadmap}
 %    \item \textbf{Results Retention:} Plans to maintain benchmark results in a publicly accessible repository with appropriate metadata and documentation, ensuring long-term usability.
 %\end{itemize}
 
-We wish to improve input/output data management by including more options 
-to save the results of \hawen, for instance this can include the hdf5 format
-which is not allowed currently.
-
+We wish to investigate the efficient of the linear algebra operations 
+when working with particularly high polynomial orders, as established
+in \cref{sec:WP1:Hawen:software}. 
+We emphasize the cost on the memory footprint for the matrix factorization,
+which is the main bottleneck in time-harmonic wave simulations using a 
+direct solver.
+We  will rely on the direct solver MUMPS to investigate the cost and benefit
+from its latest development to reduce the computational cost. However, one 
+must find a good compromise between the reduction of the memory footprint 
+(e.g., with block low-rank compression or mixed precision), and the accuracy
+of the solution.
 In~\cref{tab:WP3:Hawen:bottlenecks}, we briefly discuss the bottleneck roadmap associated to the software and relevant to the work package.
 
 \begin{table}[h!]
@@ -210,10 +217,17 @@ \subsection{12-Month Roadmap}
 
     \rowcolor{numpexgray}{\rule{0pt}{2.5ex}\color{white}\bf Bottlenecks} &  {\rule{0pt}{2.5ex}\color{white}\bf Short Description }\\ 
 
-\rowcolor{white}    B10 - Scientific Productivity & provide short description here \\
-\rowcolor{numpexlightergray}    B11 - Reproducibility and Replicability of Computation & provide short description here \\
-\rowcolor{white}    B6 - Data Management & provide short description here \\
-\rowcolor{numpexlightergray}    B7 - Exascale Algorithms & provide short description here \\
+\rowcolor{white}    B10 - Scientific Productivity & 
+Investigate the efficiency of the linear solvers with high-order polynomials
+required by HDG discretization (i.e., order higher than or equal to 7). \\
+\rowcolor{numpexlightergray}    B11 - Reproducibility and Replicability of Computation 
+& The benchmarks will be made available on online repository. \\
+\rowcolor{white}    B6 - Data Management 
+&  Inputs parameters will be made available to generate the matrix but to its size, the matrix
+   itself should not be stored.\\
+\rowcolor{numpexlightergray}    B7 - Exascale Algorithms & 
+Control the memory footprint of the solver and use latest developments of direct solver 
+MUMPS to reduce it. Investigate how the compression affects the accuracy of the solutions.\\
 \end{tabular}
         }
     }

software/hawen/WP4/WP4.tex

@@ -210,18 +210,23 @@ \subsection{12-Month Roadmap}
 We wish to combine the results of WP1 and WP3 
 \cref{sec:WP1:Hawen:software,sec:WP3:Hawen:software} 
 to obtain the most efficient configuration for inversion.
+It includes the use of high-order polynomials which, in turn,
+require to have sufficient flexibility for the representation 
+of the model onto the discretized domain to have the adequate
+resolution. In addition, the choices of configuration for the 
+underlying optimization method (e.g., choice of misfit function)
+is investigated.
 Once the numerical setup is designed, everything will be 
-made available on a dedicated repository for reproducibility.
-
+made available on a dedicated repository for reproducibility,
+see \cref{tab:WP4:Hawen:bottlenecks}.
 
 %In this section, describe the roadmap for improving benchmarks and addressing the challenges identified. This should include:
 %\begin{itemize}
 %    \item \textbf{Data Improvements:} Plans for improving input/output data management, including making datasets more accessible and ensuring reproducibility through open-data initiatives.
 %    \item \textbf{Methodology Application:} Implementation of the benchmarking methodology proposed in this deliverable to streamline reproducibility and dataset management.
 %    \item \textbf{Results Retention:} Plans to maintain benchmark results in a publicly accessible repository with appropriate metadata and documentation, ensuring long-term usability.
 %\end{itemize}
-
-In~\cref{tab:WP4:Hawen:bottlenecks}, we briefly discuss the bottleneck roadmap associated to the software and relevant to the work package.
+% In~\cref{tab:WP4:Hawen:bottlenecks}, we briefly discuss the bottleneck roadmap associated to the software and relevant to the work package.
 
 \begin{table}[h!]
     \centering
@@ -239,10 +244,18 @@ \subsection{12-Month Roadmap}
 
     \rowcolor{numpexgray}{\rule{0pt}{2.5ex}\color{white}\bf Bottlenecks} &  {\rule{0pt}{2.5ex}\color{white}\bf Short Description }\\ 
 
-\rowcolor{white}    B10 - Scientific Productivity & provide short description here \\
-\rowcolor{numpexlightergray}    B11 - Reproducibility and Replicability of Computation & provide short description here \\
-\rowcolor{white}    B6 - Data Management & provide short description here \\
-\rowcolor{numpexlightergray}    B7 - Exascale Algorithms & provide short description here \\
+\rowcolor{white}    B10 - Scientific Productivity 
+                        & Use flexible representation of the unknowns to maintain 
+                          the resolution in inversion, while still using discretization 
+                          based on large cells to benefit from the HDG method. 
+                          \\
+\rowcolor{numpexlightergray}    B11 - Reproducibility and Replicability of Computation 
+                        & Benchmarks are made available via online repositories.\\
+\rowcolor{white}    B6 - Data Management 
+                        & Datasets will be available within the repository. \\
+\rowcolor{numpexlightergray}    B7 - Exascale Algorithms 
+                        & Enable different methods to carry out the 
+                          iterative optimization.  \\
 \end{tabular}
         }
     }

software/hpddm/hpddm.tex

@@ -129,9 +129,9 @@ \subsection{Relevant Publications}
 
 Here is a list of relevant publications related to the software:
 \begin{itemize}
-\item \fullcite{jolivet_ddprecond_2013} presents a two-level domain decomposition preconditioner tested to solve elliptic problems in parallel with up to 22 billion unknowns in 2D and 2 billion unknowns in 3D.
-\item \fullcite{jolivet_blockrecycling_2016} presents block iterative methods and recycling strategies to solve linear systems with multiple right-hand sides and millions of unknwons while preserving a good scalability.
-\item \fullcite{jolivet_ksphpddm-pchpddm_2021} presents KSPHPDDM (Krylov methods) and PCHPDDM (preconditioners) available in PETSc, when configured with HPDDM support.
+\item \fullcite{jolivet_scalable_2013} presents a two-level domain decomposition preconditioner tested to solve elliptic problems in parallel with up to 22 billion unknowns in 2D and 2 billion unknowns in 3D.
+\item \fullcite{jolivet_block_2016} presents block iterative methods and recycling strategies to solve linear systems with multiple right-hand sides and millions of unknwons while preserving a good scalability.
+\item \fullcite{jolivet_ksphpddm_2021} presents KSPHPDDM (Krylov methods) and PCHPDDM (preconditioners) available in PETSc, when configured with HPDDM support.
 \end{itemize}
 
 

software/pbb/WP5/WP5.tex

@@ -128,11 +128,11 @@ \subsubsection{Benchmark \#1}
 \subsection{12-Month Roadmap}
 \label{sec:WP5:pBB:roadmap}
 
-In this section, describe the roadmap for improving benchmarks and addressing the challenges identified. This should include:
+In this section, we describe the roadmap for improving benchmarks and addressing the challenges identified. %This should include:
 \begin{itemize}
-    \item \textbf{Data Improvements:} Plans for improving input/output data management, including making datasets more accessible and ensuring reproducibility through open-data initiatives.
-    \item \textbf{Methodology Application:} Implementation of the benchmarking methodology proposed in this deliverable to streamline reproducibility and dataset management.
-    \item \textbf{Results Retention:} Plans to maintain benchmark results in a publicly accessible repository with appropriate metadata and documentation, ensuring long-term usability.
+    \item \textbf{Data Improvements:} There are no issues concerning the improvements of input/output data management. All datasets are accessible and reproducibility is ensured through open-data initiatives.
+    %\item \textbf{Methodology Application:} Implementation of the benchmarking methodology proposed in this deliverable to streamline reproducibility and dataset management.
+    \item \textbf{Results Retention:} Plans to maintain benchmark results in a publicly accessible repository with appropriate metadata and documentation, ensuring long-term usability. All obtained results are open access and are published in HAL, conferences and journals.
 \end{itemize}
 
 In~\cref{tab:WP5:pBB:bottlenecks}, we briefly discuss the bottleneck roadmap associated to the software and relevant to the work package.