• Aucun résultat trouvé

Sparse Linear Algebra Solvers

Fast and Accurate Simulation of Multithreaded Sparse Linear Algebra Solvers

Fast and Accurate Simulation of Multithreaded Sparse Linear Algebra Solvers

... each sparse matrix has a unique structure, the corresponding DAG is very different and the kernel parameters (such as height and width) greatly vary from one matrix factorization to ...single linear model ...

11

Hierarchical hybrid sparse linear solver for multicore platforms

Hierarchical hybrid sparse linear solver for multicore platforms

... Parallel sparse linear algebra solvers are often the innermost numerical kernels in scientific and engineering applications; consequently, they are one of the most time consuming ...large ...

29

Scheduling Trees of Malleable Tasks for Sparse Linear Algebra

Scheduling Trees of Malleable Tasks for Sparse Linear Algebra

... in sparse direct solvers where the matrix is first factorized before the actual solution is ...tiled linear algebra routines (BLAS): the sub-matrix is decomposed into 2D tiles (or blocks), and ...

13

Deciding Non-Compressible Blocks in Sparse Direct Solvers using Incomplete Factorization

Deciding Non-Compressible Blocks in Sparse Direct Solvers using Incomplete Factorization

... other linear algebra solvers, one of the most impor- tant operation is the block ...or sparse multifrontal solvers, the cost of the low-rank update is usually small with respect to the ...

20

Evaluation of the OGF GridRPC Data Management library, and  study of its integration into an International Sparse Linear Algebra Expert System

Evaluation of the OGF GridRPC Data Management library, and study of its integration into an International Sparse Linear Algebra Expert System

... for linear algebra aims at providing tools and softwares for sparse matrices (matrices with a higher ratio of zero components versus nonzero ...Direct Solvers currently available in GridTLSE ...

13

Implementing multifrontal sparse solvers for multicore architectures with Sequential Task Flow runtime systems

Implementing multifrontal sparse solvers for multicore architectures with Sequential Task Flow runtime systems

... Linear algebra kernels. In principle the above parallel algorithms (either 1D or 2D) could be implemented using the standard geqrt, gemqrt, tpqrt and tpmqrt LAPACK routines. This, however, would imply a ...

24

Sparse direct solvers with accelerators over DAG runtimes

Sparse direct solvers with accelerators over DAG runtimes

... The DAG U E runtime aimed first at providing a scheduler for large distributed system of multicore nodes and is able to handle heterogeneous architectures to relocate the most compute intensive kernel on GPUs. A full ...

15

On the resilience of parallel sparse hybrid solvers

On the resilience of parallel sparse hybrid solvers

... solving linear systems so large that the time between two consecutive faults may be smaller than the time required by linear algebra solvers to ...enhance sparse solvers so that ...

11

Scheduling Trees of Malleable Tasks for Sparse Linear Algebra

Scheduling Trees of Malleable Tasks for Sparse Linear Algebra

... 8 Conclusion In this paper, we have studied how to schedule trees of malleable tasks whose speedup function on multicore platforms is p α . We have first motivated the use of this model for sparse matrix ...

44

Minimax rate of testing in sparse linear regression

Minimax rate of testing in sparse linear regression

... of linear regression model under sparse ...under sparse alternatives was first studied by Ingster (1997) and Donoho and Jin (2004) who considered the Gaussian mean ...to linear regression ...

17

Identification of switched linear systems via sparse optimization

Identification of switched linear systems via sparse optimization

... of sparse optimization, ...Since sparse op- timization is in general non-convex, it is classical to con- sider instead a convex ℓ 1 relaxation of this ...

15

Dual Extrapolation for Sparse Generalized Linear Models

Dual Extrapolation for Sparse Generalized Linear Models

... Generalized Linear Models (GLM) form a wide class of regression and classification models, where prediction is a function of a linear combination of the input ...

28

Exploiting regularity in sparse Generalized Linear Models

Exploiting regularity in sparse Generalized Linear Models

... θ accel along with θ res . When Line 16 is used with this method as inner solver, we refer to it as the Newton-Celer variant. V. Experiments Implementation is done in Python and Cython [5] for the low- level critical ...

8

Stacked Sparse Non-Linear Blind Source Separation

Stacked Sparse Non-Linear Blind Source Separation

... Linear BSS: We use AMCA algorithm [4], which discards the highly non-linear samples, considering them as partial correlations (samples with sources simultaneously active). It thus finds at iteration l a ...

3

Trading Performance for Memory in Sparse Direct Solvers using Low-rank Compression

Trading Performance for Memory in Sparse Direct Solvers using Low-rank Compression

... There are few works that propose an implementation of a low-rank solver using low-rank updates. In [14], such an approach is proposed, but the per- formance obtained is low, slower than a generic sparse direct ...

32

Scheduling of Linear Algebra Kernels on Multiple Heterogeneous Resources

Scheduling of Linear Algebra Kernels on Multiple Heterogeneous Resources

... favor different task types. The main principle of H ETERO P RIO remains unchanged, though: whenever a resource is free, it picks a ready task among the task type with highest score. The first scoring system is called ...

11

Efficient algorithms and implementation in exact linear algebra

Efficient algorithms and implementation in exact linear algebra

... 1.3.3 Memory efficiency Another difficulty arising with the block Wiedemann algorithm is its memory requirement. While using a block of n vectors allows a n-fold parallelism, its memory requirement is increased by a ...

113

A Linear Algebra Framework for Static HPF Code Distribution

A Linear Algebra Framework for Static HPF Code Distribution

... These two sets cannot always be used to derive the data exchange code. Firstly, a processor p does not need to exchange data with itself. This cannot be expressed directly as an ane constraint and must be added as a ...

44

Simultaneous conversions with the Residue Number System using linear algebra

Simultaneous conversions with the Residue Number System using linear algebra

... 4. APPLICATION TO INTEGER MATRIX MULTIPLICATION An obvious application for our algorithms is integer matrix multiplication, and by extension matrix multiplication over rings of the form Z/N Z for large values of N . ...

20

A model from signal analysis to design linear algebra activities

A model from signal analysis to design linear algebra activities

... Finally, item c) intended to probe if students’ previously constructed structures about matrices and vectors enabled them to recognize the product of a matrix A and a vector s in the model for the transformation of ...

11

Show all 2730 documents...

Sujets connexes