In-Network Computing

Overview

With the slowdown of Moore’s law, large-scale data centers have been employing an increasing number of domain-specific accelerators (GPUs, TPUs, and FPGAs) to deliver the needed unprecedented performance for computation-intensive workloads like machine learning model training and inference. Under this trend, the emergence of programmable network devices (P4 switches, SmartNICs, and DPUs more recently) has motivated a new concept called in-network computing, where programmable network devices are instructed (with P4/NPL languages) to accelerate application-specific computations (e.g., AllReduce for distributed machine learning training) in addition to running network functions. In-network computing brings tremendous performance benefits for a variety of distributed workloads, but also imposes challenges to the design of data center systems. Our work on in-network computing centers around programming and management systems, aiming to facilitate the adoption in production data centers.

People

Publications

  • SoCC 2026
    FluxReduce: Straggler-Resilient In-Network Aggregation for Distributed Machine Learning
    George Karlos, Henri E. Bal, Lin Wang
    ACM Symposium on Cloud Computing (SoCC), 2026
  • ISCC 2024Second Best Paper Award
    NetNN: Neural Intrusion Detection System in Programmable Networks
    Kamran Razavi, Shayan Davari Fard, George Karlos, Vinod Nigade, Max Mühlhäuser, Lin Wang
    IEEE Symposium on Computers and Communications (ISCC), 2024
  • SC 2024
    NetCL: A Unified Programming Framework for In-Network Computing
    George Karlos, Henri E. Bal, Lin Wang
    ACM/IEEE International Conference for High Performance Computing, Networking, Storage, and Analysis (SC), 2024
  • ToN 2022
    Holistic Resource Scheduling for Data Center In-Network Computing
    Marcel Blöcher, Lin Wang, Patrick Eugster, Max Schmidt
    IEEE/ACM Transactions on Networking (ToN), 2022
  • ASPLOS 2021
    Switches for HIRE: Resource Scheduling for Data Center In-Network Computing
    Marcel Blöcher, Lin Wang, Patrick Eugster, Max Schmidt
    ACM International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS), 2021
  • HotNets 2021
    Don't You Worry 'Bout a Packet: Unified Programming for In-Network Computing
    George Karlos, Henri E. Bal, Lin Wang
    ACM Workshop on Hot Topics in Networks (HotNets), 2021