Abstract
In this paper, we present a new technique for mapping the backpropagation algorithm on hypercubes and related architectures. A key component of this technique is a network partitioning scheme called checkerboarding. Checkerboarding allows us to replace the all-to-all broadcast operation performed by the commonly used vertical network partitioning scheme, with operations that are much faster on the hypercubes and related architectures. Checkerboarding can be combined with the pattern partitioning technique to form a hybrid scheme that performs better than either one of these schemes. Theoretical analysis and experimental results on nCUBE® and CM5® show that our scheme performs better than the other schemes, for both uniform and nonuniform networks.
| Original language | English (US) |
|---|---|
| Pages (from-to) | 1073-1090 |
| Number of pages | 18 |
| Journal | IEEE Transactions on Parallel and Distributed Systems |
| Volume | 5 |
| Issue number | 10 |
| DOIs | |
| State | Published - Oct 1994 |
Bibliographical note
Funding Information:Manuscript received October 13, 1992; revised June I, 1993. This work was supported by Amy Research Office under Grant 28408-MA-SDI to the University of Minnesota, by a grant from the Graduate School of University of Minnesota, and by the Army High-Performance Computing Research Center at the University of Minnesota. The authors are with the Department of Computer Science, University of Minnesota, Minneapolis, MN 55455 USA; e-mail: [email protected], [email protected], [email protected]. IEEE Log Number 9403 102.
Keywords
- Backpropagation algorithm
- hypercube
- network partitioning
- neural networks
- pattern partitioning
- scalability