Genetic Algorithms in Search, Optimization and Machine Learning
Genetic Algorithms in Search, Optimization and Machine Learning
Interconnection Networks: An Engineering Approach
Interconnection Networks: An Engineering Approach
Networks on chip
Scheduling Collective Communications on Wormhole Fat Cubes
SBAC-PAD '05 Proceedings of the 17th International Symposium on Computer Architecture on High Performance Computing
Evolutionary design of group communication schedules for interconnection networks
ISCIS'05 Proceedings of the 20th international conference on Computer and Information Sciences
An evolutionary design technique for collective communications on optimal diameter-degree networks
Proceedings of the 10th annual conference on Genetic and evolutionary computation
Evolutionary Design of Fault Tolerant Collective Communications
ICES '08 Proceedings of the 8th international conference on Evolvable Systems: From Biology to Hardware
Evolutionary optimization of multistage interconnection networks performance
Proceedings of the 11th Annual conference on Genetic and evolutionary computation
Evolutionary-based conflict-free scheduling of collective communications on spidergon NoCs
Proceedings of the 12th annual conference on Genetic and evolutionary computation
Hi-index | 0.00 |
Since chip multiprocessors are quickly penetrating new application areas in network and media processing, their interconnection architectures become a subject of sophisticated optimization. One-to-All Broadcast (OAB) and All-to-All Broadcast (AAB) [2] group communications are frequently used in many parallel algorithms and if their overhead cost is excessive, performance degrades rapidly with a processor count. This paper deals with the design of a new application-specific standard genetic algorithm (SGA) and the use of Hybrid parallel Genetic Simulated Annealing (HGSA) to design optimal communication algorithms for an arbitrary topology of the interconnection network. Each of these algorithms is targeted for a different switching technique. The OAB and AAB communication schedules were designed mainly for an asymmetrical AMP [15] network and for the benchmark hypercube network [16] using Store-and-Forward (SF) and Wormhole (WH) switching.