Analysis of application-aware on-chip routing under traffic uncertainty

Abstract

Application-aware routing exploits static knowledge of an applica tion's traffic pattern to improve performance compared to general purpose routing algorithms. Unfortunately, traditional approaches to application-aware routing cannot efficiently handle dynamic changes in the traffic pattern limiting its usefulness in practice. In this paper, we study application-aware routing under traffic uncer tainty. Our problem formulation allows an application to statically specify an uncertainty set of traffic patterns that each occur with a given probability, and our goal is to find a single set of combined routes that will enable high-performance across all of these traffic patterns. We show how efficient combined routes can be found for this problem using convex optimization. These combined routes are optimal when the performance for every traffic pattern using the combined routes is the same as the performance using routes that are specialized for just that traffic pattern. We derive necessary and sufficient conditions for when our optimization framework will find optimal combined routes. We use theoretical and numerical analysis for the important class of permutation traffic patterns to quantify how often optimal combined routes exist and to determine the performance loss when optimal combined routes are infeasi ble. Finally, we use a cycle-level simulator that includes realistic pipeline latencies, arbitration, and buffered flow-control to study the latency and throughput of combined routes compared to spe cialized routes and routes generated using general-purpose routing algorithms. The theoretical analysis, numerical analysis, and simu lation results in this paper provide a first step towards more flexible application-aware routing.

Power-aware Manhattan Routing on Chip Multiprocessors

“…If, however, the characteristics of the traffic are statically known, then routing algorithms can take advantage of that knowledge to optimize the performance of the interconnection network. For on-chip routing, there have been many proposals to design traffic-aware routes with the goal of maximizing the communication bandwidth and/or minimizing its delay [13,8].…”

Joint Optimization of Application Specific Routing in an Anycast Network

“…To this end, the optimization carried out at the access level of network have reached to its pinnacle with multitude of optimizations ranging from MIMO, OFDMA etc. [1]. while the network router still take 1 ms to process each L3 packet.…”

“…These techniques rely upon having a centralized controller with global state information available while in real-network, the states are distributed [6], [7] and need time to converge even if dijkstra-like algorithms are used. Literature presents work relating to optimizing network with distributed states as well as joint optimization using convex methods but do not detail how to (a) implement these in live network, (b) utilize distributed states globally and (c) have a system-wide optimization [1][2][3].…”