Predictive Switching in 2-D Torus Routers
Predictive Switching in 2-D Torus Routers
复制标题
DOI：
10.1109/iwias.2006.33
复制
发表时间：
2006-01
期刊：
影响因子：
--
通讯作者：
T. Yoshinaga;S. Kamakura;M. Koibuchi
中科院分区：
文献类型：
--
作者：
T. Yoshinaga;S. Kamakura;M. Koibuchi
本文提出了预测切换在2D环面路由器，以减少低延迟通信的流水线阶段的数量。利用并行应用中的通信规律，提出了一种动态预测机制，在数据包到达路由器之前预测数据包在路由器中的路径。因此，当预测命中时，我们可以绕过路由计算、虚拟通道分配和交换机分配的流水线阶段。我们考虑了NAS并行基准测试中几种流量模式的预测器架构和准确性。我们的实验表明，采样模式匹配（SPM）预测器实现77%至96%的预测命中率时，我们使用的维序路由算法。我们还讨论了一种方法，以提高SPM的预测精度，通过检查的频率出现的预测值在通信历史
This paper proposes predictive switching in 2D torus routers to reduce the number of pipeline stages for low-latency communication. By utilizing the communication regularity in parallel applications, a dynamic predicting mechanism presets packet traversal paths inside the router before packet arrivals. Hence, we can bypass the pipeline stages of routing computation, virtual channel allocation and switch allocation when the prediction hits. We considered the predictor architecture and accuracy for several traffic patterns in NAS parallel benchmarks. Our experiments show that a sampled pattern matching (SPM) predictor achieves 77% to 96% of the prediction hit rates when we use the dimension-order routing algorithm. We also discuss a method to improve the prediction accuracy of SPM by examining the frequency of occurrence for the prediction values in the communication history