Title :
A new approach for computing multidimensional DFT´s on parallel machines and its implementation on the iPSC/860 hypercube
Author :
Kechriotis, George I. ; An, Myoung ; Bletsas, Michail ; Tolimieri, Richard ; Manolakos, Elias S.
Author_Institution :
AWARE Inc., Cambridge, MA, USA
fDate :
1/1/1995 12:00:00 AM
Abstract :
Proposes a new approach for computing multidimensional DFTs that reduces interprocessor communications and is therefore suitable for efficient implementation on a variety of multiprocessor platforms including MIMD supercomputers and clusters of workstations. Group theoretic concepts are used to formulate a flexible computational strategy that hybrids the reduced transform algorithm (RTA) with the Good-Thomas factorization and can deal efficiently with non-power-of-two sizes without resorting to zero-padding. The RTA algorithm is employed not as a data processing but rather as a bookkeeping tool in order to decompose the problem into many smaller size subproblems (lines) that can be solved independently by the processors. Implementation issues on an Intel iPSC/i860 hypercube are discussed and timing results for large 2D and 3D DFTs with index sets in Z/MP×Z/KP and Z/N×Z/MP×Z/KP respectively are provided, where N, M, K are powers-of-two and P is a small prime number such as 3, 5, or 7. The nonoptimized realizations of the new hybrid RTA approach are shown to outperform by as much as 70% the optimized assembly coded realizations of the traditional row-column method on the iPSC/860
Keywords :
discrete Fourier transforms; group theory; hypercube networks; parallel algorithms; parallel architectures; Good-Thomas factorization; Intel iPSC/i860 hypercube; MIMD supercomputers; RTA algorithm; bookkeeping tool; clusters of workstation; computational strategy; discrete Fourier transform; group theoretic concepts; implementation; interprocessor communications; multidimensional DFT; multiprocessor platforms; non-power-of-two sizes; nonoptimized realizations; parallel machines; reduced transform algorithm; subproblems; timing results; Assembly; Clustering algorithms; Concurrent computing; Data processing; Hypercubes; Multidimensional systems; Optimization methods; Supercomputers; Timing; Workstations;
Journal_Title :
Signal Processing, IEEE Transactions on