Frame Shared Memory: Line-Rate Networking on Commodity Hardware [abstract] (ACM DL, PDF)
John Giacomoni, John K. Bennet, Antonio Carzaniga, Douglas C. Sicker, Manish Vachharajani, and Alexander L. Wolf
Proceedings of the 2007 ACM/IEEE Symposium on Architectures for Networking and Communications Systems (ANCS), December 2007.
Network processors provide an economical programmable platform to
handle the high throughput and frame rates of modern and next-generation
communication systems. However, these platforms have exchanged
general-purpose capabilities for performance.
This paper presents an alternative; a software framework for commodity
general-purpose architectures capable of high-rate and throughput
sequential frame processing compatible with high-level languages
and general-purpose operating system. A cache-optimized
concurrent lock free queue provides the low-overhead core-to-core
communication necessary to sustain sequential frame processing rates
beyond the 1.41~million frames per second realized on gigabit Ethernet
while permitting per-frame processing time expansion with pipeline
parallelism.