Taurus: A Data Plane Architecture for Per-Packet ML
Authors:
Preprint
Abstract
Emerging applications -- cloud computing, the internet of things, and augmented/virtual reality -- demand responsive, secure, and scalable datacenter networks. These networks currently implement simple, per-packet, data-plane heuristics (e.g., ECMP and sketches) under a slow, millisecond-latency control plane that runs data-driven performance and security policies. However, to meet applications' service-level objectives (SLOs) in a modern data center, networks must bridge the gap between line-rate, per-packet execution and complex decision making. In this work, we present the design and implementation of Taurus, a data plane for line-rate inference. Taurus adds custom hardware based on a flexible, parallel-patterns (MapReduce) abstraction to programmable network devices, such as switches and NICs; this new hardware uses pipelined SIMD parallelism to enable per-packet MapReduce operations (e.g., inference). Our evaluation of a Taurus switch ASIC -- supporting several real-world models -- shows that Taurus operates orders of magnitude faster than a server-based control plane while increasing area by 3.8% and latency for line-rate ML models by up to 221 ns. Furthermore, our Taurus FPGA prototype achieves full model accuracy and detects two orders of magnitude more events than a state-of-the-art control-plane anomaly-detection system.Keywords
All Related Versions
- Version 1, 2020-02-12, ArXiv
- Version 2, 2022-01-19, ArXiv
- Published version: , 1099.
This publication has 86 references indexed in Scilit:
- Real World OCaml: Functional Programming for the Masses2022
- Coarse-Grained Reconfigurable Architectures2020
- What You Need to Know About SDN Flow Tables2015
- Analysis of KDD Dataset Attributes - Class wise for Intrusion DetectionProcedia Computer Science, 2015
- Support Vector Machines for TCP traffic classificationComputer Networks, 2009
- McPAD: A multiple classifier system for accurate payload-based anomaly detectionComputer Networks, 2009
- NN-RED: an AQM mechanism based on neural networksElectronics Letters, 2007
- An improved data stream summary: the count-min sketch and its applicationsJournal of Algorithms, 2005
- PipeRench: a reconfigurable architecture and compilerComputer, 2000
- Approximation capabilities of multilayer feedforward networksNeural Networks, 1991