High-Performance, Low-Complexity Deadlock Avoidance for Arbitrary Topologies/Routings

Abstract

Recently, the use of graph-based network topologies has been proposed as an alternative to traditional networks such as tori or fattrees due to their very good topological characteristics. However they pose practical implementation challenges such as the lack of deadlock avoidance strategies. Previous proposals are either exceedingly complex, underutilise network resources or lack flexibility. We propose-and prove formally-three generic, low-complexity deadlock avoidance mechanisms that only require local information. The main strengths of our method are its topology-and routingindependence and that the virtual channel count is bounded by the length of the longest path. We evaluate our proposed mechanisms against previous proposals through an extensive simulation study to measure the impact on the performance using both synthetic and realistic traffic. First we compare against a well-known HPC mechanism for dragonfly and achieved similar performance level. Then we moved to Graph-based networks and show that our mechanisms can greatly outperform traditional, spanning-tree based mechanisms, even if these use a much larger number of virtual channels. Overall, we find that our proposal provides a simple, flexible and high performance deadlock-avoidance solution.

Exaeskalarako sare-interkonexioen diseinurako helburu-aniztasuneko optimizazioa

“…INSEEk bideratzaileen eredu oso zehatza dauka (phit mailan), eta bai trafiko sintetikozko patroiak bai egiazko aplikazioetatik ateratako lan-kargak erabil ditzake. Horretaz gain, hainbat mekanismo dauzka elkar blokeatzea saihesteko topologia eta bideraketa askotarako [14]. Topologia horien errendimendua neurtzeko trafiko uniformea eta bisect izeneko trafikoa erabili ditugu.…”

“…Bestetik, goiko mailak guztiz aldakorrak dira FPGA-k erabiltzen dituzten bideratzaileak [5] darabiltzatelako. Hori dela eta, konputazio-eta sare-elementuak era batean baino gehiagotan lotu daitezke ausazko topologiak osatzeko [14]. Malgutasun horri esker bai topologia zuzenak, bideratzaileen arteko zuzeneko konexioak erabiltzen dituztenak, bai zeharkako topologiak, bideratzaile osagarriak erabiltzen dituztenak, eraiki ditzakegu.…”