A cube-connected cycles architecture with high reliability and improved performance
- 1 January 1993
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Computers
- Vol. 42 (2) , 246-253
- https://doi.org/10.1109/12.204798
Abstract
The cube-connected cycles (CCC) architecture is an attractive parallel computation network, because it is suitable for VLSI implementation while preserving all the desired features of hypercubes. However, the CCC tends to suffer from considerable performance degradation when a fault arises. In this work, a fault-tolerant CCC which exhibits significantly enhanced reliability is proposed. Reconfiguration in response to an operational fault in this fault-tolerant CCC is simple and can be performed in a distributed manner. When compared with the CCC, the proposed design in the absence of faults gets performance improvement as a result of faster broadcasting and PE-to-PE communication. The layout of this structure is discussed, and its area overhead is found to be moderate if the PE size is much larger than the link/switch size. Therefore, this design approach is particularly useful for situations where the PE is relatively complex.Keywords
This publication has 7 references indexed in Scilit:
- Yield enhancement designs for WSI cube connected cyclesPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2003
- Effects of switch failure on soft-configurable WSI yieldPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- Reconfigurable cube-connected cycles architecturesJournal of Parallel and Distributed Computing, 1990
- The cubical ring connected cycles: a fault tolerant parallel computation networkIEEE Transactions on Computers, 1988
- A normalized-area measure for VLSI layoutsIEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 1988
- Deadlock-Free Message Routing in Multiprocessor Interconnection NetworksIEEE Transactions on Computers, 1987
- The cube-connected cycles: a versatile network for parallel computationCommunications of the ACM, 1981