Gradient descent, conjugate gradient, and other iterative algorithms are a powerful class of algorithms; however, they can take a long time for conver- gence. Baseline accelerator designs feature insu cient coverage of operations and do not work well on the problems we target. In this thesis we present a novel hardware architecture for accelerating gradient descent and other similar algorithms. To support this architecture, we also present a sparse matrix-vector storage format, and software support for utilizing the format, so that it can be e ciently mapped onto hardware which is also well suited for dense operations. We show that the accelerator design outperforms similar designs which target only the most dominant operation of a given al...
AbstractIn this paper, an analysis of some of the tradeoffs involved in the design and efficient imp...
International audienceNowadays, several industrial applications are being ported to parallel archite...
This paper develops the original conjugate gradient method and the idea of preconditioning a system....
Gradient descent, conjugate gradient, and other iterative algorithms are a powerful class of algorit...
The last ten years have seen the rise of a new parallel computing paradigm with diverse hardware arc...
We characterize the performance and power attributes of the conjugate gradient (CG) sparse solver wh...
Abstract. The limiting factor for efficiency of sparse linear solvers is the memory bandwidth. In th...
© 2014 Technical University of Munich (TUM).The conjugate gradient (CG) is one of the most widely us...
Graphics Processing Units (GPUs) exhibit significantly higher peak performance than conventional CPU...
In linear solvers, like the conjugate gradient algorithm, sparse-matrix vector multiplication is an ...
In this paper, we present a sparse matrix-vector multiplication algorithm for massively-parallel com...
A decision tree is a well-known machine learning technique. Recently their popularity has increased ...
The Conjugate Gradient (CG) method is a widely-used iterative method for solving linear systems desc...
In this paper we describe a new approach for accelerating the Conjugate Gradient (CG) method using a...
Conjugate gradient is an important iterative method used for solving least squares problems. It is c...
AbstractIn this paper, an analysis of some of the tradeoffs involved in the design and efficient imp...
International audienceNowadays, several industrial applications are being ported to parallel archite...
This paper develops the original conjugate gradient method and the idea of preconditioning a system....
Gradient descent, conjugate gradient, and other iterative algorithms are a powerful class of algorit...
The last ten years have seen the rise of a new parallel computing paradigm with diverse hardware arc...
We characterize the performance and power attributes of the conjugate gradient (CG) sparse solver wh...
Abstract. The limiting factor for efficiency of sparse linear solvers is the memory bandwidth. In th...
© 2014 Technical University of Munich (TUM).The conjugate gradient (CG) is one of the most widely us...
Graphics Processing Units (GPUs) exhibit significantly higher peak performance than conventional CPU...
In linear solvers, like the conjugate gradient algorithm, sparse-matrix vector multiplication is an ...
In this paper, we present a sparse matrix-vector multiplication algorithm for massively-parallel com...
A decision tree is a well-known machine learning technique. Recently their popularity has increased ...
The Conjugate Gradient (CG) method is a widely-used iterative method for solving linear systems desc...
In this paper we describe a new approach for accelerating the Conjugate Gradient (CG) method using a...
Conjugate gradient is an important iterative method used for solving least squares problems. It is c...
AbstractIn this paper, an analysis of some of the tradeoffs involved in the design and efficient imp...
International audienceNowadays, several industrial applications are being ported to parallel archite...
This paper develops the original conjugate gradient method and the idea of preconditioning a system....