Results 11  20
of
25
A Scalable Paradigm for EffectivelyDense Matrix Formulated Applications
 Proc. of the European Conference on HighPerformance Computing and Networking
, 1994
"... ..."
(Show Context)
LAPACK Working Note 95 ScaLAPACK: A Portable Linear Algebra Library for Distributed Memory Computers  Design Issues and Performance
, 1995
"... This paper outlines the content and performance of ScaLAPACK, a collection of mathematical software for linear algebra computations on distributed memory computers. The importance of developing standards for computational and message passing interfaces is discussed. We present the different componen ..."
Abstract

Cited by 3 (0 self)
 Add to MetaCart
This paper outlines the content and performance of ScaLAPACK, a collection of mathematical software for linear algebra computations on distributed memory computers. The importance of developing standards for computational and message passing interfaces is discussed. We present the different components and building blocks of ScaLAPACK. This paper outlines the difficulties inherent in producing correct codes for networks of heterogeneous processors. We define a theoretical model of parallel computers dedicated to linear algebra applications: the Distributed Linear Algebra Machine (DLAM). This model provides a convenient framework for developing parallel algorithms and investigating their scalability, performance and programmability. Extensive performance results on various platforms are presented and analyzed with the help of the DLAM. Finally, this paper briefly describes future directions for the ScaLAPACK library and concludes by suggesting alternative approaches to mathematical libra...
Efficient use of parallel libraries on heterogeneous Networks of Workstations
, 2000
"... The paper is motivated by efficiency considerations about porting mathematical software from Massively Parallel Processors (MPPs) to Networks of Workstations (NOWs). Heterogeneity of the network is the major obstacle to efficient porting: it can be overcome with a specialized system, Programming env ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
The paper is motivated by efficiency considerations about porting mathematical software from Massively Parallel Processors (MPPs) to Networks of Workstations (NOWs). Heterogeneity of the network is the major obstacle to efficient porting: it can be overcome with a specialized system, Programming envIronment for Network of COmputers (PINCO), for monitoring available computational power at different nodes, both statically and dynamically. The structure and functionalities of PINCO are outlined, and a significant porting example, matrix multiplication, is
Numerical linear algebra algorithms and software www.elsevier.nl/locate/cam
"... The increasing availability of advancedarchitecture computers has a signi cant e ect on all spheres of scienti c computation, including algorithm research and software development in numerical linear algebra. Linear algebra – in particular, the solution of linear systems of equations – lies at the ..."
Abstract
 Add to MetaCart
(Show Context)
The increasing availability of advancedarchitecture computers has a signi cant e ect on all spheres of scienti c computation, including algorithm research and software development in numerical linear algebra. Linear algebra – in particular, the solution of linear systems of equations – lies at the heart of most calculations in scienti c computing. This paper discusses some of the recent developments in linear algebra designed to exploit these advancedarchitecture computers. We discuss two broad classes of algorithms: those for dense, and those for sparse matrices. c ○ 2000 Elsevier Science
THE DESIGN OF APARALLEL DENSE LINEAR ALGEBRA SOFTWARE LIBRARY: REDUCTION TO HESSENBERG, TRIDIAGONAL, AND BIDIAGONAL FORM
, 1995
"... Prepared by the ..."
(Show Context)
Application of Massively Parallel Architecture to Computational Electromagnetics
, 1994
"... of Dissertation In this thesis, we discuss the development and implementation of computational electromagnetics simulations on massively parallel processing systems. The possibility of predicting radar cross section (RCS) for a full scale aircraft is discussed and demonstrated by combining the most ..."
Abstract
 Add to MetaCart
of Dissertation In this thesis, we discuss the development and implementation of computational electromagnetics simulations on massively parallel processing systems. The possibility of predicting radar cross section (RCS) for a full scale aircraft is discussed and demonstrated by combining the most advanced computational electromagnetics techniques and massively parallel processing technologies. Wilkes' and Cha's exact surface model and their basis function are used to develop numerical solutions for electromagnetic scattering problems involving arbitrarily shaped conducting bodies with and without lossy dielectric coatings. The ParaMoM codeone of the most sophisticated and complicated software packages for electromagnetic scattering developed by Cha's group at Syracuse Research Corporationis extended to treat arbitrarily shaped conducting bodies with lossy dielectric coatings. The parallel algorithms development of ParaMoM is discussed. The parallel ParaMoM, called ParaMoMMPP, ...
CLUE  Cluster Evaluation
, 2000
"... This report describes the simulation tool CLUE which enables the highly accurate performance assessment and prediction of clusters of symmetric multiprocessors (SMPs). Using CLUE, reliable information can be obtained to reach the optimum decision on hardware configurations (processing elements and c ..."
Abstract
 Add to MetaCart
This report describes the simulation tool CLUE which enables the highly accurate performance assessment and prediction of clusters of symmetric multiprocessors (SMPs). Using CLUE, reliable information can be obtained to reach the optimum decision on hardware configurations (processing elements and communication networks) before actually purchasing this hardware. Thus, hardware can be adapted to individual software features, reversing the currently applied adaptation of highperformance software to hardware features (as used, for instance, in FFTW [8, 9], Phipac [2], or the Atlas tool [18]).
LAPACK Working Note 43 A Look at Scalable Dense Linear Algebra Libraries ∗
, 1992
"... We discuss the essential design features of a library of scalable software for performing dense linear algebra computations on distributed memory concurrent computers. The square block scattered decomposition is proposed as a flexible and generalpurpose way of decomposing most, if not all, dense ma ..."
Abstract
 Add to MetaCart
(Show Context)
We discuss the essential design features of a library of scalable software for performing dense linear algebra computations on distributed memory concurrent computers. The square block scattered decomposition is proposed as a flexible and generalpurpose way of decomposing most, if not all, dense matrix problems. An objectoriented interface to the library permits more portable applications to be written, and is easy to learn and use, since details of the parallel implementation are hidden from the user. Experiments on the Intel Touchstone Delta system with a prototype code that uses the square block scattered decomposition to perform LU factorization are presented and analyzed. It was found that the code was both scalable and efficient, performing at about 14 GFLOPS (double precision) for the largest problem considered.