Categorizing Network Traffic in Update-Based Protocols on Scalable Multiprocessors

In this paper we categorize the coherence traffic in update-based protocols and show that, for most applications, more than 90% of all updates generated by the protocol are unnecessary. We identify application characteristics that generate useless update traffic, and compare the isolated and combined effects of several software and hardware techniques for eliminating useless updates. These techniques include dynamic and static hybrid protocols, a data re-mapping strategy, and coalescing write buffers. Our simulations show that these techniques are effective for different types of useless updates. Overall, software caching (where dynamic data re-mapping is performed under programmer or compiler control) has the potential to significantly increase the percentage of useful traffic in applications. When software caching is not applicable, either the static or the dynamic protocol generates the least useless traffic. Although coalescing write buffers provide great reductions in the total number of messages transferred, these buffers do not necessarily increase the percentage of useful traffic.