Communication lower bounds and optimal algorithms for numerical linear algebra
نویسنده
چکیده
Acta Numerica / Volume 23 / May 2014, pp 1 155 DOI: 10.1017/S0962492914000038, Published online: 12 May 2014 Link to this article: http://journals.cambridge.org/abstract_S0962492914000038 How to cite this article: G. Ballard, E. Carson, J. Demmel, M. Hoemmen, N. Knight and O. Schwartz (2014). Communication lower bounds and optimal algorithms for numerical linear algebra . Acta Numerica, 23, pp 1-155 doi:10.1017/S0962492914000038 Request Permissions : Click here
منابع مشابه
Minimizing Communication in Numerical Linear Algebra
In 1981 Hong and Kung proved a lower bound on the amount of communication (amount of data moved between a small, fast memory and large, slow memory) needed to perform dense, n-by-n matrix-multiplication using the conventional O(n3) algorithm, where the input matrices were too large to fit in the small, fast memory. In 2004 Irony, Toledo and Tiskin gave a new proof of this result and extended it...
متن کاملCommunication-optimal Parallel and Sequential Cholesky Decomposition
Numerical algorithms have two kinds of costs: arithmetic and communication, by which we mean either moving data between levels of a memory hierarchy (in the sequential case) or over a network connecting processors (in the parallel case). Communication costs often dominate arithmetic costs, so it is of interest to design algorithms minimizing communication. In this paper we first extend known lo...
متن کاملCommunication-Optimal Parallel and Sequential Eigenvalue/SVD Algorithms
Algorithms have two costs: arithmetic and communication, by which we mean either moving data between levels of a memory hierarchy (in the sequential case) or over a network connecting processors (in the parallel case). The simplest metric of communication is to count the total number of words moved (also called the bandwidth cost). On current hardware the cost of moving a single word already gr...
متن کاملCommunication Avoiding (CA) and Other Innovative Algorithms
In 1981 Hong and Kung proved a lower bound on the amount of communication (amount of data moved between a small, fast memory and large, slow memory) needed to perform dense, n-by-n matrix-multiplication using the conventional O(n) algorithm, where the input matrices were too large to fit in the small, fast memory. In 2004 Irony, Toledo and Tiskin gave a new proof of this result and extended it ...
متن کاملTradeoffs between synchronization, communication, and work in parallel linear algebra computations
This paper derives tradeoffs between three basic costs of a parallel algorithm: synchronization, data movement, and computational cost. Our theoretical model counts the amount of work and data movement as a maximum of any execution path during the parallel computation. By considering this metric, rather than the total communication volume over the whole machine, we obtain new insight into the c...
متن کامل