Ensō: A Streaming Interface for NIC-Application Communication
Hugo Sadok, Nirav Atre, Zhipeng Zhao, Daniel S. Berger, James C. Hoe, Aurojit Panda, Justine Sherry, Ren Wang
摘要
Today, most communication between the NIC and software involves exchanging fixed-size packet buffers. This packetized interface was designed for an era when NICs implemented few offloads and software implemented the logic for translating between application data and packets. However, both NICs and networked software have evolved: modern NICs implement hardware offloads, e.g., TSO, LRO, and serialization offloads that can more efficiently translate between application data and packets. Furthermore, modern software increasingly batches network I/O to reduce overheads. These changes have led to a mismatch between the packetized interface, which assumes that the NIC and software exchange fixed-size buffers, and the features provided by modern NICs and used by modern software. This incongruence between interface and data adds software complexity and I/O overheads, which in turn limits communication performance.
This paper proposes Ensō, a new streaming NIC-to-software interface designed to better support how NICs and software interact today. At its core, Ensō eschews fixed-size buffers, and instead structures communication as a stream that can be used to send arbitrary data sizes. We show that this change reduces software overheads, reduces PCIe bandwidth requirements, and leads to fewer cache misses. These improvements allow an Ensōbased NIC to saturate a 100 Gbps link with minimum-sized packets (forwarding at 148.8 Mpps) using a single core, improve throughput for high-performance network applications by 1.5-6×, and reduce latency by up to 43%.
问问这篇 Paper
智能体会读完全文。
Lune 把这篇 Paper 索引到了每一个公式,引用它的顶会 Paper 也一样。你提问,回答直接引用原文。
引用它的顶会 Paper9
- Automatic Parallelization of Software Network FunctionsFrancisco Pereira, Fernando M. V. Ramos, Luis PedrosaNSDI 2024 · 被引用 14 次
- BBQ: A Fast and Scalable Integer Priority Queue for Hardware Packet SchedulingNirav Atre, Hugo Sadok, Justine SherryNSDI 2024 · 被引用 12 次
- High-throughput and Flexible Host Networking for Accelerated ComputingAthinagoras Skiadopoulos, Zhiqiang Xie, Mark Zhao, Qizhe Cai 等OSDI 2024 · 被引用 11 次
- CEIO: A Cache-Efficient Network I/O Architecture for NIC-CPU Data PathsBowen Liu, Xinyang Huang, Qijing Li, Zhuobin Huang 等SIGCOMM 2025 · 被引用 10 次
- SG-IOV: Socket-Granular I/O Virtualization for SmartNIC-Based Container NetworksChenxingyu Zhao, Hongtao Zhang, Jaehong Min, Shengkai Lin 等ASPLOS 2026 · 被引用 2 次
它引用的顶会 Paper19
- The CacheLib Caching Engine: Design and Experiences at ScaleBenjamin Berg, Daniel S. Berger, Sara McAllister, Isaac Grosof 等OSDI 2020 · 被引用 145 次
- Enabling Programmable Transport Protocols in High-Speed NICsMina Tahmasbi Arashloo, Alexey Lavrov, Manya Ghobadi, Jennifer Rexford 等NSDI 2020 · 被引用 96 次
- The Demikernel Datapath OS Architecture for Microsecond-scale Datacenter SystemsIrene Zhang, Amanda Raybuck, Pratyush Patel, Kirk Olynyk 等SOSP 2021 · 被引用 83 次
- From luna to solar: the evolutions of the compute-to-storage networks in Alibaba cloudRui Miao, Lingjun Zhu, Shu Ma, Kun Qian 等SIGCOMM 2022 · 被引用 79 次
- The nanoPU: A Nanosecond Network Stack for DatacentersStephen Ibanez, Alex Mallery, Serhat Arslan, Theo Jepsen 等OSDI 2021 · 被引用 74 次
相关 Paper
- Efficient Remote Memory Ordering for Non-Coherent SystemsWei Siew Liew, Md Ashfaqur Rahaman, Adarsh Patil, Ryan Stutsman 等ASPLOS 2026
- weBurst can be Harmless: Achieving Line-rate Software Traffic Shaping by Inter-flow BatchingDanfeng Shan, Shihao Hu, Yuqi Liu, Wanchun Jiang 等INFOCOM 2023 · 被引用 2 次
- BURST: Seeking High-performance, Interoperability and Scalability in Soft-RDMAHuijun Shen, Zelong Yue, Jian Yang, Zhuo Jiang 等NSDI 2026
- CC-NIC: a Cache-Coherent Interface to the NICHenry N. Schuh, Arvind Krishnamurthy, David E. Culler, Henry M. Levy 等ASPLOS 2024 · 被引用 19 次
- PANIC: A High-Performance Programmable NIC for Multi-tenant NetworksJiaxin Lin, Kiran Patel, Brent E. Stephens, Anirudh Sivaraman 等OSDI 2020 · 被引用 104 次
