DocumentCode :
1754227
Title :
NeuroNoC: Neural network inspired runtime adaptation for an on-chip communication architecture
Author :
Ebi, Thomas ; Faruque, Mohammad Abdullah Al ; Henkel, Jörg
Author_Institution :
Dept. of Embedded Syst., Karlsruhe Inst. of Technol. (KIT), Karlsruhe, Germany
fYear :
2010
fDate :
24-29 Oct. 2010
Firstpage :
223
Lastpage :
230
Abstract :
The on-chip communication architecture presented in this paper, NeuroNoC, addresses the problems arising in large multi-core systems where global or local routing strategies do not work efficiently anymore since they either do not scale or lack information on the network state. Our communication architecture is runtime adaptive and it deploys a distributed artificial neural network to aid routing decisions. It thereby provides a light-weight mechanism for local routing information to propagate through the communication architecture and is capable of self-organizing efficiently (since scalable) to varying communication workload scenarios. The underlying basic concepts are borrowed from spiking neural networks, a special case of artificial neural networks. Our experiments show that already with low hardware overhead, a significant improvement of the runtime routing behavior compared to current state-of-the-art approaches is possible. We report an improvement of 23% in routing quality compared to wXY routing in terms of failed transactions.
Keywords :
multiprocessing systems; multiprocessor interconnection networks; network-on-chip; neural nets; NeuroNoC; communication workload scenario; distributed artificial neural networks; local routing information; multicore system; neural network inspired runtime adaptation; onchip communication architecture; routing decision; spiking neural networks; Artificial neural networks; Bandwidth; Biomembranes; Computer architecture; Kernel; Neurons; Routing; Adaptive system; Network-on-Chip; Spiking Neural Networks;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Hardware/Software Codesign and System Synthesis (CODES+ISSS), 2010 IEEE/ACM/IFIP International Conference on
Conference_Location :
Scottsdale, AZ
Print_ISBN :
978-1-6055-8905-3
Type :
conf
Filename :
5751505
Link To Document :
بازگشت