Two Bilinear (3 × 3)-Matrix Multiplication Algorithms of Complexity 25


Cite item

Full Text

Open Access Open Access
Restricted Access Access granted
Restricted Access Subscription Access

Abstract

As is known, a bilinear algorithm for multiplying 3 × 3 matrices can be constructed by using ordered triples of 3 × 3 matrices Aρ, Bρ, Cρ, \(\rho = \overline {1,r} ,\) where r is the complexity of the algorithm. Algorithms with various symmetries are being extensively studied. This paper presents two algorithms of complexity 25 possessing the following two properties (symmetries): (1) the matricesA1,B1, and C1 are identity, (2) if the algorithm involves a tripleA, B, C, then it also involves the triples B, C, A and C, A, B. For example, these properties are inherent in the well-known Strassen algorithm for multiplying 2 × 2 matrices. Many existing (3 × 3)-matrix multiplication algorithms have property (2). Methods for finding new algorithms are proposed. It is shown that the found algorithms are different and new.

About the authors

B. V. Chokaev

Faculty of Computational Mathematics and Computer Technologies

Author for correspondence.
Email: g110@yandex.ru
Russian Federation, Groznyi, Chechen Republic, 364093

G. N. Shumkin

Department of Computational Mathematics and Cybernetics

Email: g110@yandex.ru
Russian Federation, Moscow, 119991

Supplementary files

Supplementary Files
Action
1. JATS XML

Copyright (c) 2018 Allerton Press, Inc.