🤖 AI Summary
This work addresses the inherent limitation of static on-chip network routing in achieving effective load balancing due to its inability to adapt to runtime traffic conditions. To overcome this, the authors introduce N-Rank—a novel approach that models long-term traffic load trends determined jointly by network topology and traffic patterns—and integrates it with BiDOR, a lightweight runtime routing mechanism that dynamically selects paths based on these trends. This hybrid strategy preserves the simplicity and predictability of static routing while enabling quasi-static dynamic optimization. Experimental results demonstrate significant performance gains: under uniform traffic, throughput improves by 42.9%, and under real application workloads, average and maximum packet latencies are reduced by 86.4% and 95.3%, respectively.
📝 Abstract
In networks-on-chip, static routing schemes are favored for their simplicity and predictability, but they cannot effectively balance network load due to the unawareness of runtime load distribution. Q-StaR discovers two factors (topology and traffic distribution) that determine the long-term trend of load distribution, and proposes N-Rank to extract this trend. The obtained information is used to guide BiDOR's route selection at runtime, thereby improving load balancing while retaining simplicity and predictability. Simulation validates that Q-StaR significantly outperforms the typical dimension-order routing (throughput under uniform traffic improved by 42.9\%, and mean/maximum latency under realistic workloads reduced by 86.4\%/95.3\%).