In-network Computing with SmartNICs for HPC Applications
Supercomputing Conference Tutorial
Chicago, Illinois
November 15, 2026
Overview
The past few years have witnessed an increased level of support for and deployment of programmable network adapters, known as "SmartNICs." These enhanced network devices offer standard packet processing capabilities as well as advanced "in-network" computing features built around programmable lightweight processing cores, FPGAs, and even CPU- and GPU-based platforms capable of running separate operating systems. SmartNICs have gained rapid adoption for data center tasks, including infrastructure management, packet filtering, and I/O acceleration. Increasingly these devices are also being explored for high-performance computing (HPC) and AI application acceleration. This tutorial provides an in-depth exploration of the state of the art in SmartNICs and the emerging software ecosystems that support them. Attendees will participate in hands-on exercises designed to demonstrate how SmartNICs can be leveraged to accelerate HPC and AI applications. Topics include packet processing frameworks such as P4, I/O acceleration techniques, MPI offloading, and algorithmic adaptations for utilizing SmartNIC processors. Participants will have the opportunity to run these exercises on cutting-edge platforms, including NVIDIA BlueField-3 Data Processing Units (DPUs), using cloud-based environments such as USC’s NETLAB and the FABRIC testbed. The tutorial will also highlight additional strategies for optimizing applications to effectively use SmartNICs as communication accelerators in HPC systems.
Audience
This tutorial is intended for HPC users, application developers, researchers and developers of programming models and communication libraries, as well as tool developers who are interested in leveraging next-generation SmartNICs for HPC.
Tutorial Goals
By participating in this comprehensive tutorial, attendees will gain:
- An understanding of asynchronous programmable engines, such as SmartNICs, and their evolution in HPC architectures, including an overview of current efforts by major vendors such as NVIDIA, Intel, and AMD.
- Familiarity with programming models for SmartNICs, such as vendor-supported frameworks like P4 and DOCA, OpenMP offloading, and communication offloading with MPI.
- Practical knowledge of leveraging SmartNICs for in-line packet processing, communication offload optimizations, storage optimizations, and algorithmic changes in applications.
- Real-world application experiences and mini-apps case studies that leverage SmartNICs and DPUs
- Hands-on experience with exercises covering a variety of application examples, including tutorials on P4 and DOCA features, blocking and nonblocking MPI collective offload operations, OpenMP offload for DPUs, and using accelerators like Data Path Accelerators (DPAs).
Pre-requisites
Connectivity to the Internet and a browser to access the online virtual platform. Attendees will be provided with an account to access USC’s NETLAB system: https://netlab.cec.sc.edu/
Agenda
Sunday, November 15
| Time (GMT) | Topic | Presenter |
|---|---|---|
| 8:30-8:40 | Introduction | Elie Kfoury (USC), Jeffrey Young (Gatech), Jorge Crichigno (USC), Richard Graham (NVIDIA), Antonio Peña (BSC), Aaron Jezghani (Gatech) |
| 8:40-8:55 | Communication Offloading | Elie Kfoury (USC), Jeffrey Young (Gatech), Jorge Crichigno (USC), Richard Graham (NVIDIA), Antonio Peña (BSC), Aaron Jezghani (Gatech) |
| 8:55-9:15 | SmartNIC Use Cases | Elie Kfoury (USC), Jeffrey Young (Gatech), Jorge Crichigno (USC), Richard Graham (NVIDIA), Antonio Peña (BSC), Aaron Jezghani (Gatech) |
| 9:15-10:00 | Infrastructure Software – DOCA and P4 Introduction | Elie Kfoury (USC), Jeffrey Young (Gatech), Jorge Crichigno (USC), Richard Graham (NVIDIA), Antonio Peña (BSC), Aaron Jezghani (Gatech) |
| 10:00-10:30 | BREAK | |
| 10:30-11:30 | Hands-on DOCA/P4 | Elie Kfoury (USC), Jeffrey Young (Gatech), Jorge Crichigno (USC), Richard Graham (NVIDIA), Antonio Peña (BSC), Aaron Jezghani (Gatech) |
| 11:30-12:00 | HPC Programming – Part 1 | Elie Kfoury (USC), Jeffrey Young (Gatech), Jorge Crichigno (USC), Richard Graham (NVIDIA), Antonio Peña (BSC), Aaron Jezghani (Gatech) |
| 12:00-1:30 | LUNCH | |
| 1:30-2:15 | Hands-on with DPA Programming | Elie Kfoury (USC), Jeffrey Young (Gatech), Jorge Crichigno (USC), Richard Graham (NVIDIA), Antonio Peña (BSC), Aaron Jezghani (Gatech) |
| 2:15-2:45 | HPC Programming – Part 2 | Elie Kfoury (USC), Jeffrey Young (Gatech), Jorge Crichigno (USC), Richard Graham (NVIDIA), Antonio Peña (BSC), Aaron Jezghani (Gatech) |
| 2:45-3:00 | Hands-on HPC Demo – OpenMP/MPI Offload with ODOS | Elie Kfoury (USC), Jeffrey Young (Gatech), Jorge Crichigno (USC), Richard Graham (NVIDIA), Antonio Peña (BSC), Aaron Jezghani (Gatech) |
| 3:00-3:30 | BREAK | |
| 3:30-3:40 | Tutorial Survey | |
| 3:40-4:00 | Storage Acceleration | Elie Kfoury (USC), Jeffrey Young (Gatech), Jorge Crichigno (USC), Richard Graham (NVIDIA), Antonio Peña (BSC), Aaron Jezghani (Gatech) |
| 4:00-5:00 | Continued Hands-on Activities | Elie Kfoury (USC), Jeffrey Young (Gatech), Jorge Crichigno (USC), Richard Graham (NVIDIA), Antonio Peña (BSC), Aaron Jezghani (Gatech) |
| Slides [PPT, PDF] | ||