Re-architecting datacenter networks and stacks for low latency and high performance

By Mark James Handley, Costin Raiciu, Alexandru Agache, A. Voinescu, Andrew William Moore, Gianni Antichi, M. Wójcik. Conference on Applications, Technologies, Architectures, and Protocols for Computer Communication, 2017

447 citations recorded by the source corpus.

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

A link layer protocol for quantum networks (2019, background): “for SDN [16], data center networks [43], content delivery networks [22] or cloud”
Swift: Delay is Simple and Effective for Congestion Control in the Datacenter (2020, methodology): “When number of flows exceed network BDP, Credit-based: pHost [21], NDP [23], Homa [39] and Express-Pass [15], rely on the receiver end-host issuing credit packets.”
Homa: a receiver-driven low-latency transport protocol using network priorities (2018, methodology): “The state-of-the-art approach, NDP [15], strictly limits queues to 8 packets, equivalent to roughly 10 μs of latency at 10 Gbps.”
HPCC: high precision congestion control (2019, background): “…[42] needs a new switch architecture for lossless fabric and performs per-packet adaptive load balancing of packet routes in switches; HOMA [33] and NDP [22] are receiver-driven, credit-based solutions, which is a big shift from the state-of-the-art in practice, since they have complex receivers.”