Theor. Comput. Sci. 410 (2009), 2649–2658 (DOI).
arXiv preprint (October 2008).
We describe a cache-friendly version of van der Hoeven's truncated FFT and inverse truncated FFT, focusing on the case of ‘large’ coefficients, such as those arising in the Schönhage–Strassen algorithm for multiplication in Z[x]. We describe two implementations and examine their performance.