No description available.

A preconditioner to improve the convergence properties of Krylov subspace solvers is derived and analyzed in this work. This method is adapted to linear systems arising from a finite-element discretization of a phase-field crystal equation.

Cette thèse traite de la diffusion par des surfaces rugueuses monodimensionnelles. Les surfaces présentant des petites échelles de variations nécessitent une discrétisation fine pour représenter les effets de diffusion sur le champ diffracté, ce qui augmente les coûts numériques. Deux aspects sont considérés : la réduction de la taille du problème en construisant une condition aux limiteséquivalente traduisant les effets des variations rapides et la réduction du nombre d’itérations nécessaires pour résoudre le système linéaire issu de la méthode des moments par une méthode basée sur les sous-espaces de Krylov. En ce qui concerne la réduction de la taille du problème, une technique d’homogénéisation est utilisée pour transformer la condition aux limites posée sur lasurface rugueuse par des paramètres effectifs. Ces paramètres sont déterminés par des problèmes auxiliaires qui tiennent compte des échelles fines de la surface. Dans le cas de surfaces parfaitement métalliques, la procédure est appliquée en polarisation Transverse Magnétique (TM) et Transverse Électrique (TE). Une impédance équivalente de Léontovich d’ordre 1 est déduite.Le procédure est automatique et les ordres supérieurs sont dérivés pour la polarisation TM. La procédure d’homogénéisation est aussi appliquée pour des interfaces rugueuses séparant deux milieux diélectriques. En ce qui concerne la réduction du nombre d’itérations, un préconditionneur, basé sur des considérations physiques, est construit à partir des modes de Floquet. Bien que le préconditionneur soit initialement élaboré pour des surfaces périodiques, nous montrons qu’il est aussi efficace pour des surfaces tronquées éclairées par une onde plane. L’efficacité des deux aspects présentés dans cette thèse est numériquement illustrée pour des configurations d’intérêt. / This work is about the scattering by monodimensional rough surfaces. Surfaces presenting small scales of variations need a very refined mesh to finally capture the scattering field behaviour what increases the computational cost. Two aspects are considered : the reduction of the problemsize through an effective boundary condition incorporating the effect of rapid variations and the reduction of the number of iterations to solve the linear system arising from method of moments by a method based on Krylov subspace. Firstly, an homogenization process is used to convert the boundary condition on the rough interface into effective parameters. These parameters are determined by the solutions of auxiliary problems which involve the detailed profile of the interface. In the case of perfectly metallic surfaces, the process is applied to the E- and H-polarization and an Leontovich impedance of order 1 is deduced. The process is automatic and higher orders are derived for E-polarization. The homogenization process is also applied to dielectric rough interfaces. Secondly, a physically-based preconditioner is built with Floquet’s modes. Although the preconditioner has been designed for periodical surfaces, it was shown to be efficient in the case of truncated surfaces illuminated by a plane wave. The efficiency of both aspects is numerically illustrated for some configurations of interest.

The limiting behavior of weighted paths associated with the semidefinite program (SDP) map $X^{1/2}SX^{1/2}$ was studied and some applications to error bound analysis and superlinear convergence of a class of primal-dual interior-point methods were provided. A new approach for solving large-scale well-structured sparse SDPs via a saddle point mirror-prox algorithm with ${cal O}(epsilon^{-1})$ efficiency was developed based on exploiting sparsity structure and reformulating SDPs into smooth convex-concave saddle point problems. An iterative solver-based long-step primal-dual infeasible path-following algorithm for convex quadratic programming (CQP) was developed. The search directions of this algorithm were computed by means of a preconditioned iterative linear solver. A uniform bound, depending only on the CQP data, on the number of iterations performed by a preconditioned iterative linear solver was established. A polynomial bound on the number of iterations of this algorithm was also obtained. One efficient ``nearly exact' type of method for solving large-scale ``low-rank' trust region subproblems was proposed by completely avoiding the computations of Cholesky or partial Cholesky factorizations. A computational study of this method was also provided by applying it to solve some large-scale nonlinear programming problems.

Over the last 25 years, interior-point methods (IPMs) have emerged as a viable class of algorithms for solving various forms of conic optimization problems. Most IPMs use a modified Newton method to determine the search direction at each iteration. The system of equations corresponding to the modified Newton system can often be reduced to the so-called normal equation, a system of equations whose matrix ADA' is positive definite, yet often ill-conditioned. In this thesis, we first investigate the theoretical properties of the maximum weight basis (MWB) preconditioner, and show that when applied to a matrix of the form ADA', where D is positive definite and diagonal, the MWB preconditioner yields a preconditioned matrix whose condition number is uniformly bounded by a constant depending only on A. Next, we incorporate the results regarding the MWB preconditioner into infeasible, long-step, primal-dual, path-following algorithms for linear programming (LP) and convex quadratic programming (CQP). In both LP and CQP, we show that the number of iterative solver iterations of the algorithms can be uniformly bounded by n and a condition number of A, while the algorithmic iterations of the IPMs can be polynomially bounded by n and the logarithm of the desired accuracy. We also expand the scope of the LP and CQP algorithms to incorporate a family of preconditioners, of which MWB is a member, to determine an approximate solution to the normal equation. For the remainder of the thesis, we develop a new preconditioning strategy for solving systems of equations whose associated matrix is positive definite but ill-conditioned. Our so-called adaptive preconditioning strategy allows one to change the preconditioner during the course of the conjugate gradient (CG) algorithm by post-multiplying the current preconditioner by a simple matrix, consisting of the identity matrix plus a rank-one update. Our resulting algorithm, the Adaptive Preconditioned CG (APCG) algorithm, is shown to have polynomial convergence properties. Numerical tests are conducted to compare a variant of the APCG algorithm with the CG algorithm on various matrices.

Diese Arbeit behandelt die effiziente Vorkonditionierung von Finite-Elemente-Matrizen unter Verwendung hierarchischer Matrizen.

The objective of this PhD research programme is to investigate the effectiveness of a finite volume method incorporating radial basis functions for simulating nonlinear transport processes. The finite volume method is the favoured numerical technique for solving the advection-diffusion equations that arise in transport simulation. The method transforms the original problem into a system of nonlinear, algebraic equations through the process of discretisation. The accuracy of this discretisation determines to a large extent the accuracy of the final solution. A new method of discretisation is presented that employs radial basis functions (rbfs) as a means of local interpolation. When combined with Gaussian quadrature integration methods, the resulting finite volume discretisation leads to accurate numerical solutions without the need for very fine meshes, and the additional overheads they entail. The resulting nonlinear, algebraic system is solved efficiently using a Jacobian-free Newton-Krylov method. By employing the new method as an extension of existing shape function-based approaches, the number of nonlinear iterations required to obtain convergence can be reduced. Furthermore, information obtained from these iterations can be used to increase the efficiency of subsequent rbf-based iterations, as well as to construct an effective parallel reconditioner to further reduce the number of nonlinear iterations required. Results are presented that demonstrate the improved accuracy offered by the new method when applied to several test problems. By successively refining the meshes, it is also possible to demonstrate the increased order of the new method, when compared to a traditional shape function basedmethod. Comparing the resources required for both methods reveals that the new approach can be many times more efficient at producing a solution of a given accuracy.

Cette thèse traite d’une nouvelle classe de préconditionneurs qui ont pour but d’accélérer la résolution des grands systèmes creux, courant dans les problèmes scientifiques ou industriels, par les méthodes itératives préconditionnées. Pour appliquer ces préconditionneurs, la matrice d’entrée doit être réorganisée avec un algorithme de dissection emboîtée. Nous introduisons également une technique de recouvrement qui s’adapte à l’idée de chevauchement des sous-domaines provenant des méthodes de décomposition de domaine, aux méthodes de dissection emboîtée pour améliorer la convergence de nos préconditionneurs.Les résultats montrent que cette technique de recouvrement nous permet d’améliorer la vitesse de convergence de Nested SSOR (NSSOR) et Nested Modified incomplete LU with Rowsum proprety (NMILUR) qui sont des préconditionneurs que nous étudions. La dernière partie de cette thèse portera sur nos contributions dans le domaine du calcul parallèle. Nous présenterons la distribution des données et les algorithmes parallèles utilisés pour la mise en oeuvre de nos préconditionneurs. Les résultats montrent que sur une grille régulière 400x400x400, le nombre d’itérations nécessaire à la résolution avec un de nos préconditionneurs, Nested Filtering Factorization préconditionneur (NFF), n’augmente que légèrement quand le nombre de sous-domaines augmente jusqu’à 2048. En ce qui concerne les performances d’exécution sur le super-calculateur Curie, il passe à l’échelle jusqu’à 2048 coeurs et il est 2,6 fois plus rapide que le préconditionneur Schwarz Additif Restreint (RAS) qui est un des préconditionneurs basés sur les méthodes de décomposition de domaine implémentés dans la bibliothèque de calcul scientifique PETSc, bien connue de la communauté. / This thesis addresses a new class of preconditioners which aims at accelerating solving large sparse systems arising in scientific and engineering problem by using preconditioned iterative methods. To apply these preconditioners, the input matrix needs to be reordered with K-way nested dissection. We also introduce an overlapping technique that adapts the idea of overlapping subdomains from domain decomposition methods to nested dissection based methods to improve the convergence of these preconditioners. Results show that such overlapping technique improves the convergence rate of Nested SSOR (NSSOR) and Nested Modified Incomplete LU with Rowsum property (NMILUR) precondtioners that we worked on. We also present the data distribution and parallel algorithms for implementing these preconditioners. Results show that on a 400x400x400 regular grid, the number of iterations with Nested Filtering Factorization preconditioner (NFF) increases slightly while increasing the number of subdomains up to 2048. In terms of runtime performance on Curie supercomputer, it scales up to 2048 cores and it is 2.6 times faster than the domain decomposition preconditioner Restricted Additive Schwarz (RAS) as implemented in PETSc.

<p>The broad implementation of three-dimensional full-core modeling, with pin-resolved detail, for computational simulation and analysis of nuclear reactors highlights the importance of accuracy and efficiency in simulation codes for accurate and precise analysis. The primary objective of this dissertation is to develop a high-fidelity code capable of solving time-dependent neutron transport problems with 3D whole-core pin-resolved detail in nuclear reactor cores. Additionally, the dissertation explores the optimization of the code's parallelism to enhance its computational efficiency. To reduce the computational intensity associated with the direct 3D calculation of the neutron transport equation, a high-fidelity neutron transport code called PANDAS-MOC is developed using the 2D/1D approach. The 2D radial solution is obtained using the 2D Method of Characteristics (MOC), the axial 1D solution is determined through the Nodal Expansion Method (NEM), and then two solutions are coupled using transverse leakages to find the 3D solution. The convergence of the iterative scheme is accelerated using the multi-level coarse finite different mesh (ML-CMFD) technique. The code's validation and verification are carried out using the C5G7-TD benchmark exercises.</p> <p><br></p> <p>The significant and innovative aspect of this work involves parallelizing and optimizing the PANDAS-MOC code. Three parallel models are developed and evaluated based on the distributed memory and shared memory architecture: MPI parallel model (PMPI), Segment OpenMP threading hybrid model (SGP), and Whole-code OpenMP threading hybrid model (WCP). When computing the steady state of the C5G7 3D core with the same resources, the obtained speedup relationship between the three models is PMPI \(>\) WCP \(>\) SGP, whereas the WCP model only consumed 60\% of the memory of the PMPI model. Furthermore, the hybrid reduction in the ML-CMFD solver and the parallelism design of the MOC sweep are significant issues that decreased the speedup of WCP. Therefore, this study also addresses further optimizations of these two modules.</p> <p><br></p> <p>Concerning the MOC parallelism, two improvements are discussed: No-atomic schedule and Additional Axial Decomposition (AAD) parallelism. The No-atomic schedule evenly distributed the workload among threads and removes the \textit{omp atomic} clause from the code by predefining the MOC calculation sequence for each launched OpenMP thread while ensuring a thread-safe parallel environment. It can significantly reduce the calculation time and improve parallel efficiency. Furthermore, AAD divides the axial layers and OpenMP threads into multiple groups and restricts each thread to work on the layers designated to the same group. </p> <p>Meanwhile, Flag-Save-Update reduction is designed to increase the computational efficiency of the hybrid MPI/OpenMP reduction operations in the ML-CMFD module. It is accomplished by using the global arrays and status flags and establishing a tree configuration of all threads, and it includes no implicit and explicit barriers. In the case of the C5G7 3D core, the parallel efficiency of the MOC solver is about 0.872 when using 32 threads (=\#MPI \(\times\)\#OpenMP), and the Flag-Save-Update reduction yielded better speedup than the traditional hybrid MPI/OpenMP reduction, and its superiority is more obvious as more OpenMP threads are utilized. As a result, the WCP model outperforms the PMPI model for the overall steady-state calculation.</p> <p><br></p> <p>This research also investigates parallelizable preconditioners to accelerate the convergence of the generalized minimal residual method (GMRES) in the CMFD solver. Preconditioners such as Incomplete LU factorization (ILU), Symmetric Successive Over-relaxation (SOR), and Reduced Symmetric Successive Over-Relaxation (RSOR), are implemented in PANDAS-MOC. Except for RSOR, others are unsuitable for hybrid MPI/OpenMP parallel machines due to their inherent sequential nature and dependency on computation order. Their counterparts using the Red-Black ordering algorithm, namely RB-SOR, RB-RSOR, and RB-ILU, are formatted and examined on benchmark reactors such as TWIGL-2D, C5G7-2D, C5G7-3D, and their corresponding subplane models (TWIGL-2D(5S), C5G7-2D(5S), C5G7-3D(5S)), with relaxed convergence criteria (\(10^{-3}\)). Results show that all preconditioners significantly reduce the required number of iterations to converge the GMRES solutions, and RB-SOR is the best one for most reactors. In the case of C5G7-3D(5S), preconditioners exhibit similar sublinear speedup but demonstrate varying runtimes across all tests for both MG-GMRES and 1G-GMRES. However, the speedup results in 1G-GMRES are more than twice as high as those in MG-GMRES. RB-RSOR has an optimal efficiency of 0.6967 at (4,8), while RB-SOR and RB-ILU have optimal efficiencies of 0.6855 and 0.7275 at (32,1), respectively.</p>

Cette thèse étudie les méthodes de décomposition de domaine généralement classées soit comme des méthodes de Schwarz avec recouvrement ou des méthodes par sous-structuration s'appuyant sur des sous-domaines sans recouvrement. Nous nous focalisons principalement sur la méthode des éléments finis joints, aussi appelée la méthode mortar, une approche non conforme des méthodes par sous-structuration impliquant des contraintes de continuité faible sur l'espace d'approximation. Nous introduisons un framework élément fini pour la conception et l'analyse des préconditionneurs par sous-structuration pour une résolution efficace du système linéaire provenant d'une telle méthode de discrétisation. Une attention particulière est accordée à la construction du préconditionneur grille grossière, notamment la principale variante proposée dans ce travailutilisant la méthode de Galerkin Discontinue avec pénalisation intérieure comme problème grossier. D'autres méthodes de décomposition de domaine, telles que les méthodes de Schwarz et la méthode dite three-field sont étudiées dans l'objectif d'établir un environnement de programmation générique d'enseignement et de recherche pour une large gamme de ces méthodes. Nous développons un framework de calcul avancé et dédié à la mise en oeuvre parallèle des méthodesnumériques et des préconditionneurs introduits dans cette thèse. L'efficacité et la scalabilité des préconditionneurs, ainsi que la performance des algorithmes parallèles sont illustrées par des expériences numériques effectuées sur des architectures parallèles à très grande échelle. / This thesis investigates domain decomposition methods, commonly classified as either overlapping Schwarz methods or iterative substructuring methods relying on nonoverlapping subdomains. We mainly focus on the mortar finite element method, a nonconforming approach of substructuring method involving weak continuity constraints on the approximation space. We introduce a finiteelement framework for the design and the analysis of the substructuring preconditioners for an efficient solution of the linear system arising from such a discretization method. Particular consideration is given to the construction of the coarse grid preconditioner, specifically the main variantproposed in this work, using a Discontinuous Galerkin interior penalty method as coarse problem. Other domain decomposition methods, such as Schwarz methods and the so-called three-field method are surveyed with the purpose of establishing a generic teaching and research programming environment for a wide range of these methods. We develop an advanced computational framework dedicated to the parallel implementation of numerical methods and preconditioners introduced in this thesis. The efficiency and the scalability of the preconditioners, and the performance of parallel algorithms are illustrated by numerical experiments performed on large scale parallel architectures.