scientific article; zbMATH DE number 3997125

From MaRDI portal
Publication:4723256





[For the entire collection see Zbl 0608.00015.] An algorithm for the parallel operation of the fast Fourier transform (FFT) for a MIMD system is presented. With reasonable memory requirements, the algorithm attempts to avoid the repeated calculations of trigonometric functions and the frequent accesses to critical regions in the process of re-ordering the points. The program has been tested on the NEPTUNE system and the results and analysis of timings are presented. This algorithm can be executed by any number of processors, but generally the number is much less than the N data. In the case that the number of processors is a power of two, an even more effective algorithm can be designed.




Cited in
(48)








This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4723256)