ShuttleNoC: Power-Adaptable Communication Infrastructure for Many-Core Processors
ShuttleNoC: Power-Adaptable Communication Infrastructure for Many-Core Processors
复制标题
DOI：
10.1109/tcad.2018.2855165
复制
发表时间：
2019-08
影响因子：
2.9
通讯作者：
Hang Lu;Yisong Chang;Guihai Yan;Ning Lin;Xin Wei;Xiaowei Li
中科院分区：
文献类型：
--
作者：
Hang Lu;Yisong Chang;Guihai Yan;Ning Lin;Xin Wei;Xiaowei Li
片上网络（Networks-on-chip，NoCs）作为众核处理器的通信基础设施，随着技术规模的扩大，其功耗沿着增大。然而，由于片上业务的时间和空间异质性，一个关键问题是，NoC功耗不能有效地适应其业务强度的变化，也称为局部功率自适应，因此产生次优的功率效率。现有的方法要么求助于过度供应的片上网络设计或粗粒度的带宽缩放，以部分地减轻由业务的时间或空间异构性带来的过度功耗。而在本文中，我们提出了一种新的NoC架构称为穿梭NoC（ShuttleNoC），以解决这一挑战。它利用链路重新配置来实现多个子网之间的灵活数据包遍历，并利用专门的妙语来加速延迟敏感的流量。在专用功率自适应机制的支持下，评估表明，所提出的ShuttleNoC架构可以有效地解决功率和性能的权衡，并与最先进的基线相比显着提高功率效率。
Networks-on-chip (NoCs), as the communication infrastructure in many-core processors, has demonstrated remarkable power consumption along with the technology scaling. However, due to the temporal and spatial heterogeneity of the on-chip traffic, one critical problem is that the NoC power consumption cannot effectively adapt to the variation of its traffic intensity, also known as localized power adaptation, hence yielding a suboptimal power efficiency. Prior approaches either resort to the over-provisioned NoC design or coarse-grained bandwidth scaling to partially alleviate excessive power consumption brought by the traffic temporal or spatial heterogeneity. While in this paper, we propose a novel NoC architecture called Shuttle NoC (ShuttleNoC) to address this challenge. It leverages the link reconfiguration to enable flexible packet traversing between multiple subnetworks, and specialized punch lines to accelerate latency sensitive traffic. With the support of the dedicated power adaptation mechanisms, it is shown in the evaluation that the proposed ShuttleNoC architecture could effectively tackle the power and performance tradeoff and significantly boost the power efficiency compared with the state-of-the-art baselines.