parallel FFT

The lecture started out with a slide presentation, regarding the upcoming final project presentations. Follow the links for the pdf file of the slides or to download the LaTeX source:

After recalling the significance of Fourier transforms, the algebraic importance of the discrete Fourier transform, and the O(n log(n)) running time of the fast Fourier transform, we introduced the isoefficiency function as a third tool to evaluate the performance of parallel algorithms. We derived the isoefficiency function for the parallel FFT, showing its scalability.

Bibliography