FlexDriver: a network driver for your accelerator
Haggai Eran, Maxim Fudim, Gabi Malka, Gal Shalom, Noam Cohen, Amit Hermony, Dotan Levi, Liran Liss, Mark Silberstein
Abstract
We propose a new system design for connecting hardware and FPGA accelerators to the network, allowing the accelerator to directly control commodity Network Interface Cards (NICs) without using the CPU. This enables us to solve the key challenge of leveraging existing NIC hardware offloads such as virtualization, tunneling, and RDMA for accelerator networking. Our approach supports a diverse set of use cases, from direct network access for disaggregated accelerators to inline-acceleration of the network stack, all without the complex networking logic in the accelerator.
To demonstrate the feasibility of this approach, we build Flex-Driver (FLD), an on-accelerator hardware module that implements a NIC data-plane driver. Our main technical contribution is a mechanism that compresses the NIC control structures by two orders of magnitude, allowing FLD to achieve high networking scalability with low die area cost and no bandwidth interference with the accelerator logic.
The prototype for NVIDIA Innova-2 FPGA SmartNICs showcases our design's utility for three different accelerators: a disaggregated LTE cipher, an IP-defragmentation inline accelerator, and an IoT cryptographic-token authentication offload. These accelerators reach 25 Gbps line rate and leverage the NIC for RDMA processing, VXLAN tunneling, and traffic shaping without CPU involvement.
Ask about this paper
Your agent reads all of it.
Lune indexed this paper to the last equation, along with the top-tier papers that cite it. Ask a question and the answer quotes them.
Your agent calls
Luneget_paper_fulltext
Free to start. No credit card required.
Terminal
Install the CLIlune papers fulltext d5698c0d-5086-405f-8e32-f774d93f6b6cCited by top-tier papers7
- Waverunner: An Elegant Approach to Hardware Acceleration of State Machine ReplicationMohammadreza Alimadadi, Hieu Mai, Shenghsun Cho, Michael Ferdman et al.NSDI 2023 · 16 citations
- OSMOSIS: Enabling Multi-Tenancy in Datacenter SmartNICsMikhail Khalilov, Marcin Chrapek, Siyuan Shen, Alessandro Vezzu et al.USENIX ATC 2024 · 14 citations
- ACCL+: an FPGA-Based Collective Engine for Distributed ApplicationsZhenhao He, Dario Korolija, Yu Zhu, Benjamin Ramhorst et al.OSDI 2024 · 12 citations
- Data Motion Acceleration: Chaining Cross-Domain Multi AcceleratorsShu-Ting Wang, Hanyang Xu, Amin Mamandipoor, Rohan Mahapatra et al.HPCA 2024 · 10 citations
- RpcNIC: Enabling Efficient Datacenter RPC Offloading on PCIe-attached SmartNICsJie Zhang, Hongjing Huang, Xuzheng Chen, Xiang Li et al.HPCA 2025 · 6 citations
Builds on9
- Do OS abstractions make sense on FPGAs?Dario Korolija, Timothy Roscoe, Gustavo AlonsoOSDI 2020 · 114 citations
- PANIC: A High-Performance Programmable NIC for Multi-tenant NetworksJiaxin Lin, Kiran Patel, Brent E. Stephens, Anirudh Sivaraman et al.OSDI 2020 · 104 citations
- Reexamining Direct Cache Access to Optimize I/O Intensive Applications for Multi-hundred-gigabit NetworksAlireza Farshin, Amir Roozbeh, Gerald Q. Maguire Jr., Dejan KosticUSENIX ATC 2020 · 88 citations
- StRoM: smart remote memoryDavid Sidler, Zeke Wang, Monica Chiosa, Amit Kulkarni et al.EuroSys 2020 · 83 citations
- 1RMA: Re-envisioning Remote Memory Access for Multi-tenant DatacentersArjun Singhvi, Aditya Akella, Dan Gibson, Thomas F. Wenisch et al.SIGCOMM 2020 · 70 citations
Related papers
- Lynx: A SmartNIC-driven Accelerator-centric Architecture for Network ServersMaroun Tork, Lina Maudlej, Mark SilbersteinASPLOS 2020 · 64 citations
- SmartNS: Enabling Line-rate and Flexible Network Stack with SmartNICXuzheng Chen, Jie Zhang, Baolin Zhu, Xueying Zhu et al.EuroSys 2026
- Analyzing Near-Network Hardware Acceleration with Co-Processing on DPUsDimitrios Giouroukis, Dwi P. A. Nugroho, Varun Pandey, Steffen Zeuch et al.VLDB 2025 · 3 citations
- Conspirator: SmartNIC-Aided Control Plane for Distributed ML WorkloadsYunming Xiao, Diman Zad Tootaghaj, Aditya Dhakal, Lianjie Cao et al.USENIX ATC 2024 · 13 citations
- FpgaNIC: An FPGA-based Versatile 100Gb SmartNIC for GPUsZeke Wang, Hongjing Huang, Jie Zhang, Fei Wu et al.USENIX ATC 2022 · 58 citations
