We present a deflated version of the conjugate gradient algorithm for solving linear systems. The new algorithm can be useful in cases when a small number of eigenvalues of the iteration matrix are very close to the origin. It can also be useful when solving linear systems with multiple right-hand sides, since the eigenvalue information gathered from solving one linear system can be recycled for solving the next systems and then updated.
1 January 2000 in SIAM Journal on Matrix Analysis and Applications
Many scientific applications require one to solve successively linear systems Ax=b with different right-hand sides b and a symmetric positive definite matrix A. The conjugate gradient method applied to the first system generates a Krylov subspace which can be efficiently recycled thanks to orthogonal projections in subsequent systems. A modified conjugate gradient method is then applied with a specific initial guess and initial descent direction and a modified descent direction during the iterations. This paper gives new theoretical results for this method and proposes a new version. Numerical experiments show the efficacy of our method even for quite different right-hand sides.
1 January 2013 in Computing in Science & Engineering
To reduce the design complexity of OpenCL programming, the approach proposed here generates application code automatically, based on model-driven engineering (MDE) and modeling and analysis of real-time and embedded (MARTE) systems. The aim is to provide application-development resources for nonspecialists in parallel programming, exploiting concepts such as reuse and platform independence.
Optimization using finite element analysis and the adjoint variable method to solve engineering problems appears in various application areas. However, to the best of the authors’ knowledge, there is a lack of detailed explanation on the implementation of the adjoint variable method in the context of electromagnetic modeling. This paper aimed to provide a detailed explanation of the method in the simplest possible general framework. Then, an extended explanation is offered in the context of electromagnetism. A discrete design methodology based on adjoint variables for magnetostatics was formulated, implemented, and verified. This comprehensive methodology supports both linear and nonlinear problems. The framework provides a general approach for performing a very efficient and discretely consistent sensitivity analysis for problems involving geometric and physical variables or any combination of the two. The accuracy of the implementation is demonstrated by independent verification based on an analytical test case and using the finite-difference method. The methodology was used to optimize the parameters of a superconducting energy storage device and a magnet press and the optimization of the topology of an electromagnet. The objective function of each problem was successfully decreased, and all constraints stipulated were met.
1 November 2008 in Applied Mathematics and Computation
The electrical and electronic engineerings have used parallel programming to solve their large scale complex problems for performance reasons. However, as parallel programming requires a non-trivial distribution of tasks and data, developers find it hard to implement their applications effectively. Thus, in order to reduce design complexity, we propose an approach to generate code for hybrid architectures (e.g., CPU + GPU) using OpenCL, an open standard for parallel programming of heterogeneous systems. This approach is based on Model Driven Engineering (MDE) and the MARTE profile, standard proposed by Object Management Group (OMG). The aim is to provide resources to non-specialists in parallel programming to implement their applications. Moreover, thanks to model reuse ability, we can add/change functionalities and the target architecture. Consequently, this approach helps industries to achieve their time-to-market constraints which are confirmed here by experimental tests. Besides the software development at high-level abstractions, this approach aims to improve performance by using multi-GPU environments. A case study based on the Conjugate Gradient method gives clarity to our methodology.