Conference Paper

Power and Area Optimization of 3D Networks-on-Chip Using Smart and Efficient Vertical Channels.

Conference: Integrated Circuit and System Design. Power and Timing Modeling, Optimization, and Simulation - 21st International Workshop, PATMOS 2011, Madrid, Spain, September 26-29, 2011. Proceedings
Source: DBLP

ABSTRACT 3D NoC offers greater device integration, faster vertical interconnects and more power efficient inter-layer communication due to the beneficial attribute of short through silicon via (TSV) in 3D IC technologies. However, TSV pads used for bonding to a wafer layer, occupy significant chip area and result in routing congestions and expensive manufacturing process. This can lead to a significant reduction in 3D ICs' yield and higher power densities compared to 2D NoCs. In this paper, a power-efficient and low-cost inter-layer communication scheme is proposed as one way to mitigate these challenges. Instead of using a pair of unidirectional channels for inter-layer communication, utilizing a high-performance bidirectional channel enables a system to benefit from low-latency nature of the vertical interconnects and to remarkably reduce the number of TSVs. Additionally, we present a forecasting-based dynamic frequency scaling technique for reducing the power consumption of the inter-layer communication. Our extensive simulations demonstrate significant area and power improvements compared to a typical symmetric 3D NoC.

  • [Show abstract] [Hide abstract]
    ABSTRACT: The emerging three-dimensional integrated circuits (3D-ICs) achieve greater device integration and enhanced system performance at lower cost and reduced area footprint, thereby offering higher order of connectivity and greater design choices and possibilities. To exploit the intrinsic capability of reduced communication distances in 3D-ICs, three-dimensional NoC-bus hybrid mesh architecture was proposed. Besides its various advantages in terms of area, power consumption, and performance, this architecture has a unique and hitherto previously unexplored possibility to implement an efficient system-wide monitoring network. In this paper, an efficient three-dimensional NoC architecture is proposed which is optimized for system performance, power consumption, and reliability. The mechanism benefits from a congestion-aware and bus failure-tolerant routing algorithm called AdaptiveZ for vertical communication. In addition, we have integrated a low-cost monitoring platform on top of the three-dimensional NoC-Bus Hybrid mesh architecture that can be efficiently used for various system management purposes such as traffic monitoring, fault tolerance, and thermal management. The proposed generic monitoring platform called ARB-NET utilizes bus arbiters to exchange the monitoring information directly with each other without using the data network. As a test case, based on the proposed monitoring platform, a fully congestion-aware and interlayer fault-tolerant routing algorithm named AdaptiveXYZ is presented taking advantage of information generated within bus arbiters. Compared to recently proposed stacked mesh three-dimensional NoCs, our extensive simulations with synthetic and real benchmarks reveal that our architecture using the AdaptiveXYZ routing can help in achieving significant power, performance, and reliability improvements with a negligible hardware overhead.
    IEEE Transactions on Computers 03/2014; 63(3):734-747. DOI:10.1109/TC.2012.278 · 1.47 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: Three-dimensional integrated circuits (3D ICs) offer greater device integration, reduced signal delay and reduced interconnect power. It also provides greater design flexibility by allowing heterogeneous integration. However, 3D technology exacerbates the on-chip thermal issues and increases packaging and cooling costs. In order to exploit the intrinsic capability of reducing the wire length in 3D ICs, 3D NoC-Bus Hybrid mesh architecture was proposed. This architecture provides a seemingly significant platform to implement an integrated low-cost system-wide monitoring network. In this paper, a generic monitoring and management platform called ARB-NET is presented. Based on the ARB-NET monitoring platform, a fully congestion-aware adaptive routing algorithm named AdaptiveXYZ is provided which takes advantage from viable information generated within the monitoring network. In addition, we address both the power and thermal issues of a stacked mesh 3D network on chips using AdaptiveXYZ routing. To this end, a thermal model of a 3D stacked NoC system in a modern flip-chip package is developed. Thermal and power analysis are performed in order to investigate the impact of the proposed adaptive routing from the power and thermal perspectives. Our experiments with a videoconference encoder as a real application show significant power, performance and peak temperature improvements compared to a typical stacked mesh 3D NoC.
    Digital System Design (DSD), 2012 15th Euromicro Conference on; 01/2012
  • [Show abstract] [Hide abstract]
    ABSTRACT: Through Silicon Via (TSV) is the state-of-the-art vertical interconnect technology in three dimensional Integrated Circuits (3D-ICs). TSVs offer short wire length with low capacitive load and, hence, fast connections between two or more chip layers. On the other hand, TSVs consume a relative large amount of chip area and are error prone during manufacturing resulting in a dramatic yield drop for large TSV counts. Because of their short wire length, TSVs can be clocked much higher than conventional intra-layer links. To efficiently utilize the vertical bandwidth of TSVs, this paper proposes multiplexing several virtual links with dynamically allocated bit rates for guaranteed service connections via a shared TSV-Hub-Array. Virtual links can be state-of-the-art interconnects like busses, crossbars or 2D-NoC links. The TSV-Hub allows migration of traditional 2D interconnects towards the 3D stack while benefiting from a reduced TSV count and reuse of existing IP blocks and interconnection schemes. Furthermore, the TSV-Hub approach is also advantageous under interconnect resilience considerations. An incorporated switchbox enables dynamic protection switching for several faulty TSVs. Moreover, it can even cope with situations when more than the number of spare TSVs becomes defective. By means of a case study with two independent AXI interconnects, we could show an area reduction in the range of at least 10% for a TSV size of 10 μm and conservatively estimated the reliability improvement by one order of magnitude in comparison to a direct link interconnection.
    Microprocessors and Microsystems 11/2013; 37(8):823–835. DOI:10.1016/j.micpro.2013.04.013 · 0.60 Impact Factor