This repo includes snippets for various Programmable Transport Protocols that are deployed on hardware for extremely low latency in data centers. Ideally, for the bounded packet processing latency, the snippets would run on a NIC/CPU architecture such as nanoPU [1] to achieve high performance.
[1] Stephen Ibanez, Alex Mallery, Serhat Arslan, Theo Jepsen, Muhammad Shahbaz, Nick McKeown, and Changhoon Kim. "The nanoPU: Redesigning the CPU-Network Interface to Minimize RPC Tail Latency." Arxiv.org (2020).