Gray codes, fast Fourier transforms and hypercubes
From MaRDI portal
The usefulness of the hypercube architecture is stressed for the implementation of the fast Fourier transform. It is so because the data communications at every step of computation are only needed between neighbouring processors. If the data are distributed according to a binary reflected Gray code, the computing time increases a little as an effect due to imbalance in the load between processors.
Recommendations
Cited in
(12)- FFT algorithms and their adaptation to parallel processing
- Resource allocation strategies for hypercube architectures
- Product-shuffle networks: Toward reconciling shuffles and butterflies
- On constructing multiple spanning trees in a hypercube
- An efficient algorithm for gray-to-binary permutation on hypercubes
- Fault-tolerant algorithm for Fast Fourier Transform on hypercubes
- Gray codes for reflection groups
- A new parallel strategy for two-dimensional incompressible flow simulations using pseudo-spectral methods
- Optimal embeddings of butterfly-like graphs in the hypercube
- Efficient Computation of Middle Levels Gray Codes
- Pseudospectral methods on massively parallel computers
- The parallel Fourier pseudospectral method
This page was built for publication: Gray codes, fast Fourier transforms and hypercubes
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1097676)