In-network machine learning using programmable network devices: A survey
Machine learning is widely used to solve networking challenges, ranging from traffic
classification and anomaly detection to network configuration. However, machine learning …
classification and anomaly detection to network configuration. However, machine learning …
A comprehensive survey on smartnics: Architectures, development models, applications, and research directions
The end of Moore's Law and Dennard Scaling has slowed processor improvements in the
past decade. While multi-core processors have improved performance, they are limited by …
past decade. While multi-core processors have improved performance, they are limited by …
Unleashing SmartNIC packet processing performance in P4
SmartNICs are on the rise as a packet processing platform, with the trend towards a uniform
P4 programming model. However, unleashing SmartNIC packet processing performance in …
P4 programming model. However, unleashing SmartNIC packet processing performance in …
Isolation mechanisms for {High-Speed}{Packet-Processing} pipelines
Data-plane programmability is now mainstream. As we find more use cases, deployments
need to be able to run multiple packet-processing modules in a single device. These are …
need to be able to run multiple packet-processing modules in a single device. These are …
Sifter: An {Inversion-Free} and {Large-Capacity} Programmable Packet Scheduler
Packet schedulers play a crucial role in determining the order in which packets are served.
They achieve this by assigning a rank to each packet and sorting them based on these …
They achieve this by assigning a rank to each packet and sorting them based on these …
A vision for runtime programmable networks
Our community has made significant progress in developing programmable network
infrastructure, starting from the control plane and expanding to the data plane. As a latest …
infrastructure, starting from the control plane and expanding to the data plane. As a latest …
Disaggregating stateful network functions
D Bansal, G DeGrace, R Tewari, M Zygmunt… - … USENIX Symposium on …, 2023 - usenix.org
For security, isolation, metering and other purposes, public clouds today implement complex
network functions at every server. Today's implementations, in software or on FPGAs and …
network functions at every server. Today's implementations, in software or on FPGAs and …
Leo: Online {ML-based} Traffic Classification at {Multi-Terabit} Line Rate
Online traffic classification enables critical applications such as network intrusion detection
and prevention, providing Quality-of-Service, and real-time IoT analytics. However, with …
and prevention, providing Quality-of-Service, and real-time IoT analytics. However, with …
Poster: Continual Network Learning
We make a case for in-network Continual Learning as a solution for seamless adaptation to
evolving network conditions without forgetting past experiences. We propose implementing …
evolving network conditions without forgetting past experiences. We propose implementing …
Multitenant {In-Network} Acceleration with {SwitchVM}
S Khashab, A Rashelbach, M Silberstein - 21st USENIX Symposium on …, 2024 - usenix.org
We propose a practical approach to implementing multitenancy on programmable network
switches to make in-network acceleration accessible to cloud users. We introduce a Switch …
switches to make in-network acceleration accessible to cloud users. We introduce a Switch …