Network-on-Chip Implementation Research Articles

Accurate Model for Network-on-Chip Performance Evaluation Based on Timed Colored Petri Net

Network-on-Chip (NoC) is a power architecture that emerged to solve communication issues present in modern Systems-on-Chip (SoCs). NoC based architectures are very scalable and offer high levels of communication parallelism, among other features. Every efficient NoC implementation requires several design steps to accomplish indices of performance. Although there are many system level models, high-level models for NoC are representative in the context of design since they provide fast and accurate analysis, with low modeling effort, for further VHDL implementations. This work proposes a NoC model based on a Timed Colored Petri Net (TCPN) that computes performance indices seamlessly. Network latency and buffer occupation are of special interest in our approach as they represent the key indices when assessing NoC performance. As results, we have validated and refined the model of a 5×5 mesh NoC comparing its indices with equivalent VHDL RTL description under synthetic and real traffic situations. The proposed model is capable of analyzing the influence of the router service time on the average latency time, enabling internal NoC evaluation to optimize buffer length. Simulation results demonstrate the model suitability for latency evaluation with time estimation errors often below 1%. Furthermore, this paper discusses the effort required to extend the model with other NoC architectural features. We conclude that the use of a TCPN model of NoC generates accurate results providing as much detailed information as their equivalent experiments using VHDL description.

ILP formulation and heuristic method for energy-aware application mapping on 3D-NoCs

The rapid increase in the number of cores on chips forced the designers to invent new communication methods such as Network-on-Chip (NoC) paradigm. Advances in integrated circuit fabrications even allowed three-dimensional NoC (3D-NoC) implementations. 3D-NoCs have more advantages than their 2D counterparts such as lower area, higher throughput, better performance, and less energy consumption. However, they lack the design automation algorithms. An important design problem for a given application is mapping it on a 3D-NoC topology. In this paper, we present an integer linear programming (ILP) formulation and a novel heuristic algorithm, called CastNet3D, for application mapping onto mesh-based 3D-NoCs with energy minimization being the objective. The algorithm tries to utilize vertical links for communicating nodes as much as possible. Vertical links are shorter than horizontal ones; therefore, they are faster and consume less energy. We compared CastNet3D against ILP in terms of energy consumption and execution time on several benchmarks. Our results show that CastNet3D obtains close to optimum results in much shorter time frames.

AsyncBTree: Revisiting Binary Tree Topology for Efficient FPGA-Based NoC Implementation

Binary tree topology generally fails to attract network on chip (NoC) implementations due to its low bisection bandwidth. Fat trees are proposed to alleviate this issue by using increasingly thicker links to connect switches towards the root node. This scheme is very efficient in interconnected networks such as computer networks, which use generic switches for interconnection. In an NoC context, especially for field programmable gate arrays (FPGAs), fat trees require more complex switches as we move higher in the hierarchy. This restricts the maximum clock frequency at which the network operates and offsets the higher bandwidth achieved through using fatter links. In this paper, we discuss the implementation of a binary tree-based NoC, which achieves better bandwidth by varying the clock frequency between the switches as we move higher in the hierarchy. This scheme enables using simpler switch architecture, thus supporting higher maximum frequency of operation. The effect on bandwidth and resource requirement of this architecture is compared with other FPGA-based NoCs for different network sizes and traffic patterns.

Network-on-Chip Multicast Architectures Using Hybrid Wire and Surface-Wave Interconnects

The network-on-chip (NoC) has been introduced as an efficient communication backbone to tackle the increasing challenges of on-chip communication. Nevertheless, merely metal-based NoC implementation offers only limited performance and power scalability in terms of multicast and broadcast traffics. To meet scalability demands, this paper addresses the system-level challenges for intra-chip multicast communication in a proposed hybrid interconnects architecture. This hybrid NoC combines and utilizes both regular metal on-chip interconnects and new type of wireless-NoC (WiNoC) which is Zenneck surface wave interconnects (SWI). Moreover, this paper embeds novel multicast routing and arbitration schemes to address system-level multicast-challenges in the proposed architecture. Specifically, a design exploration of contention handling in SWI layer is considered in both centralized and decentralized manners. Consequently, the hybrid wire-SWI architecture avoids overloading the network, alleviates the formation of traffic hotspots and avoid deadlocks that are typically associated with state-of-the-art multicast handling. The evaluation is based on a cycle-accurate simulation and hardware description. It demonstrates the effectiveness of the proposed architecture in terms of power consumption (up to around $10$ x) and performance (around $22$ x) compared to regular NoCs. These results are achieved with negligible hardware overheads. This study explore promising potential of the proposed architecture for current and future NoC-based many-core processors.

Experimental evaluation and comparison of two recent Network-on-Chip routers for FPGAs

Rapid growth in the number of Intellectual Property (IP) cores in System-on-Chip (SoC) resulted in the need for effective and scalable interconnect scheme for system components – Network-on-Chip (NoC). Router is a key component in an NoC design that impacts the overall area utilization. It is crucial to evaluate the area efficiency of NoC routers. In this paper, we evaluate and compare two recent NoC routers for Field Programmable Gated Arrays (FPGAs). The first one is generated using the automated NoC synthesis tool CONfigurable NEtwork Creation Tool (CONNECT). The second one is an NoC router manually designed using VHDL and synthesized Altera Quartus II CAD tool. Three NoC topologies namely ring, mesh and torus are used for evaluating the two routers based on area utilization metric. The routers are evaluated by varying the node sizes from 4 to 16 for each topology. For smaller NoC topologies, CONNECT router uses less area but as the NoC size increases manual router design provides up to 85% reduction in area utilization. The results presented in this paper will be useful to designers interested in NoC implementation on FPGAs.

BlackOut: Enabling fine-grained power gating of buffers in Network-on-Chip routers

The Network-on-Chip (NoC) router buffers play an instrumental role in the performance of both the interconnection fabric and the entire multi-/many-core system. Nevertheless, the buffers also constitute the major leakage power consumers in NoC implementations. Traditionally, they are designed to accommodate worst-case traffic scenarios, so they tend to remain idle, or under-utilized, for extended periods of time. The under-utilization of these valuable resources is exemplified when one profiles real application workloads; the generated traffic is bursty in nature, whereby high traffic periods are sporadic and infrequent, in general. The mitigation of the leakage power consumption of NoC buffers via power gating has been explored in the literature, both at coarse (router-level) and fine (buffer-level) granularities. However, power gating at the router granularity is suitable only for low and medium traffic conditions, where the routers have enough opportunities to be powered down. Under high traffic, the sleeping potential rapidly diminishes. Moreover, disabling an entire router greatly affects the NoC functionality and the network connectivity. This article presents BlackOut, a fine-grained power-gating methodology targeting individual router buffers. The goal is to minimize leakage power consumption, without adversely impacting the system performance. The proposed framework is agnostic of the routing algorithm and the network topology, and it is applicable to any router micro-architecture. Evaluation results obtained using both synthetic traffic patterns and real applications in 64-core systems indicate energy savings of up to 70%, as compared to a baseline NoC, with a near-negligible performance overhead of around 2%. BlackOut is also shown to significantly outperform–by 35%, on average–two current state-of-the-art power-gating solutions, in terms of energy savings.

IP Protection of Mesh NoCs Using Square Spiral Routing

Intellectual property (IP) core reuse is essential for the design process of system-on-chip (SoC). Network-on-chip (NoC) has been used as an independent IP core during SoC design. However, the NoC has not been protected via IP protection and paid attention on its innovations. This paper proposes the first known approach to protect the authorship and the usage legitimacy of NoCs using specially designed routing, square spiral routing. The special routing algorithm exploits routing redundancy inherent in the mesh NoCs and transports packets along the paths, which have very low probability to be taken under commonly used routing algorithms. These unique and diverse paths are exploited in this paper to embed information of the author and identify the legal buyer of NoCs, showing high robustness and credibility. The hardware implementation of an IP-protected mesh NoC shows that the area overhead is small, which is $\sim 0.74$ %, and the power overhead is $\sim 0.52$ %, while the functionality and performance of the network is not affected. In this paper, the approach is presented for the mesh NoC, but the idea is equally applicable to other NoC topologies where the unique and diverse paths also inherently exist.

Architecture and Network-on-Chip Implementation of a New Hierarchical Interconnection Network

A Midimew-connected Mesh Network (MMN) is a minimal distance mesh with wrap-around links network of multiple basic modules (BMs), in which the BMs are 2D-mesh networks that are hierarchically interconnected for higher-level networks. In this paper, we present the architecture of the MMN, addressing of node, routing of message, and evaluate the static network performance of MMN, TESH, mesh and torus networks. In addition, we propose the network-on-chip (NoC) implementation of MMN. With innovative combination of diagonal and hierarchical structure, the MMN possesses several attractive features, including constant degree, small diameter, low cost, small average distance, moderate bisection width and high fault tolerant performance than that of other conventional and hierarchical interconnection networks. The simple architecture of MMN is also highly suitable for NoC implementation. To implement all the links of level-3 MMN, only four layers are needed which is feasible with current and future VLSI technologies.

Modeling and evaluation of ring-based interconnects for Network-on-Chip

A popular network topology for Network-on-Chip (NoC) implementations is the two-dimensional mesh, which has its drawbacks in the communication latency scalability, and the concentration of the traffic in the center of the mesh. In this paper, we consider the addition of simple and hierarchical rings to the mesh network. We propose several composite topologies that use the ring networks to reduce hop counts and latencies of global (long distance) traffic. Furthermore, we study two alternative ring architectures. The first is a slotted ring architecture that is suitable for NoC implementation due to its simplicity and low cost. The second uses wormhole routing and virtual channels, providing increased flexibility and better performance. Simulation results show that the composite architectures decrease the latencies and hop counts incurred by global traffic, thereby validating our claim that the use of hierarchical rings for global routing can in fact increase the scalability of the normal mesh network used for NoC implementations.

A Node Encoding of Torus Topology and Its Improved Routing Algorithm

With the feature size of semiconductor technology reducing and intellectual properties (IP) cores increasing, on chip communication architectures have a great influence on the performance and area of System-on-Chip (SoC) design. Network-on-Chip (NoC) has been proposed as a promising solution to complex SoC communication problems and has been widely accepted by academe and industry. This paper discusses how to choose suitable topology and node encoding scheme for NoC, and proposes a two-dimensional plane code based on Johnson Code by the combination of Torus topology with corresponding node encoding. The node encoding implies the relation among neighbouring nodes and the global information of routing. And it has good scalable characteristics. The two methods for code compressing are also presented to reduce the storage space of node address and increase the utilization rate of channel bandwidth. Utilizing the code, the improved X-Y routing is proposed, which is implemented with only three or six logic operations in middle nodes. The node structure is designed at the same time. The experimental results show combination of the proposed code with Torus topology can simplify the routing algorithm in the implementation of NoC, decrease silicon resource consumption and greatly improve communication performance.