Skip to main content

efficient 2D router architecture to extend 3D NoC-based multi-core architecture performance

Opoku Agyeman, M. and Zong, W. (2016) An efficient 2D router architecture for extending the performance of inhomogeneous 3D NoC-based multi-core architectures. In: SBAC-PAD Workshop on Applications for Multi-Core Architectures. USA: IEEE . 

To meet the performance and scalability demands of the fast-paced technological growth towards exascale and Big-Data processing with the performance bottleneck of conventional metal based interconnects, alternative interconnect fabrics such as inhomogeneous three dimensional integrated Network-on-Chip (3D NoC) has emanated as a cost-effective solution for emerging multi-core design. However, these interconnects trade-off optimized performance for cost by restricting the number of area and power hungry 3D routers. Consequently, in this paper, we propose a low-latency adaptive router with a low-complexity single-cycle bypassing mechanism to alleviate the performance degradation due to the slow 2D routers in inhomogeneous 3D NoCs. By combining the low-complexity bypassing technique with adaptive routing, the proposed router is able to balance the traffic in the network to reduce the average packet latency under various traffic loads. Simulation shows that, the proposed router can reduce the average packet delay by an average of 45% in 3D NoCs.

To read more go or

All views and opinions are the author's and do not necessarily reflected those of any organisation they are associated with. Twitter: @scottturneruon

Popular posts from this blog

Experiments in teaching Neural Networks

Excel Based

More details available at including links to the code.

All views and opinions are the author's and do not necessarily reflected those of any organisation they are associated with. Twitter: @scottturneruon

Social Analysis of Publications

The Computing staff's network of co-authors, at the University of Northampton, based on the University's  research repository NECTAR - on 12th November 2016. The data goes back to 2010.

The data was analysed using the software VOSviewer - free software for visualising networks. Differences in colours represents, the clusters of publications with those authors picked out by the software. The relative size of the circles is the relative number of publications listed; so for the two biggest circles/hubs it relates to 55 and 34 publications in this time period. Some relatively new authors, to the University but not to research, explains some of the 'islands' and the number of publications within it - it only reflects publications whilst at the University of Northampton.

To dig a little deeper, going to  look at the two biggest 'hubs' through their NECTAR records, so potentially going …

Computer lecturer’s research helps improve the next generation of technology

Taken from: A computing lecturer at the University of Northampton, who is researching into how the efficiency of our everyday devices, such as mobile phones, can be improved, has been awarded the best paper at two recent conferences. Dr Michael Opoku Agyeman has written several journal papers focusing on how the next generation of technology can meet the ever increasing demands from consumers. He was invited to present his work at the 19th Euromicro Conference on Digital System Design in Cyprus and the Institute of Electrical and Electronics Engineers’ 14th International Conference in Paris. Part of his research concentrates on whether several processing elements can be incorporated on a single chip, known as System-On-Chip, to improve the efficiency and speed of the computing systems that we use every day, from mobile phones to video-game consoles and even medical equipment…