Current research in interfacing clusters within Hierarchical Networks-on-Chip (HNoC) as well as interfacing NoC-based systems adopts a centralized approach. In this approach, a specific Processing Element (PE) acts as a gateway between interfacing peripherals and the rest of NoC elements. This paper evaluates this approach and show that it is not optimal for handling the inter-NoC communication. Routing inter-NoC traffic through a system to its gateway PE deteriorates the network performance. Results show that both the throughput and latency of the centralized approach degrade with the increase in the inter-NoC traffic bandwidth. To alleviate this, we propose a novel distributed approach, which separates the inter-NoC traffic from the intra-NoC one. Our approach relies on distributed buffers to allow PEs to efficiently communicate with the interfacing peripheral. We evaluate our approach against other interfacing ones using synthetic traffic as well as real benchmark applications. Our evaluation covers the whole system performance as well as its inter-and intra-NoC parts. Results prove that the proposed approach outperforms previous interfacing ones in terms of throughput and latency. The proposed approach significantly enhances the inter-NoC performance without any deterioration in the intra-NoC one. Considering the inter-NoC performance, we achieve a throughput that is close to the maximum possibly attainable one. Other approaches show major performance degradation, reaching as low as 10% of this maximum attainable throughput.
1 July 2021 in Computers and Electrical Engineering
Network delay is a major design parameter for Networks-on-Chip (NoC)-based applications. Improving NoC delay could be achieved at different design phases. At the system level, we study in this paper the impact of the network topology on NoC system delay using graph-theoretic concepts. A topology-based model is developed to calculate the average NoC delay, which is caused by links and routers. The proposed model could be used to select the optimal topology that achieves the minimum network delay for a given NoC application. A case study is presented to show how this model could be used to improve the delay of a given NoC application at early design phases.
2 July 2021 in International Transactions in Operational Research
In this paper, we present a novel approach in Networks-on-Chip topology optimization, by considering the network power consumption, packet transmission delay, and system reliability, simultaneously. We use the Particle Swarm Optimization technique to acquire the most suitable topology architecture, which achieves maximum reliability as well as minimum delay and power consumption. The optimization problem, which considers six design variables: network topology architecture, traffic distribution, processing elements' mapping, noise power, voltage swing, and probability of edge failure, is validated through a case study of an H.263-encoder MP3-decoder.