Abstract—In this paper, we introduce a matrix multiplication algorithm on the new born parallel architecture named Centralized Diamond architecture. Matrix multiplication is a fundamental mathematical operation which has numerous applications in most of scientific fields. There are some sorting and searching algorithms on Centralized Diamond architecture which have appropriate time complexity. Once different kinds of parallel algorithms are applicable to a parallel architecture, the architecture is considered to be more functional and can be employed for different purposes and applications. The theoretical analysis proves that the matrix multiplication operation on Centralized Diamond architecture runs in a competitive number of steps compared to existing solutions.

Index Terms—Centralized diamond architecture, EREW, matrix multiplication, parallel processing.

The authors are with the Department of Computer Science, Firoozkooh Branch, Islamic Azad University, Firoozkooh, Iran.

Cite:Masumeh Damrudi, Kamal Jadidy Aval, "Parallel Matrix Multiplication on Centralized Diamond Architecture," International Journal of Computer and Communication Engineering vol. 5, no. 2, pp. 151-157, 2016.