AI networking is the combined market for the switches, network interface cards, optics, cables and software that move training and inference data between accelerators. It excludes the servers and the accelerators themselves. As of October 2026, it is the fastest-growing slice of data center capital spending, and its size determines how much Ethernet, optics and interconnect capacity the industry must build each year.
How Big Is It Today?
It depends on what you count. Count only AI back-end Ethernet switching and NICs, and the 800G generation alone should exceed $50B across 2026 and 2027, with nearly 200M ports shipped in that window, as of July 2026. Add front-end and storage networks, and the addressable base roughly doubles. Add optics, cabling and scale-up interconnect, and the total climbs further. We track each of these layers separately, and this page explains what each bucket contains and where the published numbers sit.
We published in June 2024 that data center AI networking would surge to nearly $20B in 2025, led by Ethernet, InfiniBand and 800G optical transceivers (see the press release). The result landed: Ethernet AI networking revenue surged over 150% Y/Y in 2024, with bandwidth up over 200% Y/Y. We now project data center AI networking to surge to over $25B in 2028 (see the press release). The pace is without precedent: the industry now ships more networking capacity in a single quarter than it shipped in the entire prior year.
How Big Does It Get by 2030?
We project AI scale-out Ethernet revenue will surge past $100B by 2030, and pluggable optics at 800G and 1.6T plus emerging co-packaged optics could exceed 50% of annual switch revenue by 2030. Our published long-range view is that AI networking approaches and potentially exceeds $1T annually in the early 2030s (see The $1 Trillion Dollar Networking Market). The telling datapoint is cadence rather than any single speed: more than 200 ports a minute will barely keep pace with AI demand.
Which Segments Grow Fastest?
The 800G upgrade cycle is the heart of the AI factory build-out. We project, as of July 2026, that the 800G market for Ethernet Switching and NICs will exceed $50B in 26/27 with nearly 200M ports shipped during that time (see Ethernet Switching Is the Heart of AI Factories). Switch silicon stair-steps underneath it, from 25.6 Tbps parts in 2022 to 51.2 Tbps parts in 2024 and 2025, with 102.4 Tbps platforms emerging in 2H 2026 (see 51.2 Tbps to Drive Record-Setting 800 Gbps Volume). Port speeds themselves ramp from 400G to 800G with 1.6T on the horizon (see What Are 400G, 800G and 1.6T Ports?).
Optics are the second engine. 800G optical transceivers led the AI networking surge, and coherent pluggables are carrying the build-out beyond the data center wall (see Coherent Lite market to top 10 million units). The wider story of how capital spending flows into networking sits in Navigating the Explosion in Data Center Networking Demand.
How Was This Calculated?
The figures in this Explainer come from 650 Group’s published quarterly research on data center switching, AI networking, and interconnect markets, which track vendor revenue and port shipments by speed each quarter. Each figure above links to the publication it came from, and the full catalog is on our research reports page.
What to read next: the wider stack primer in What Is Data Center Networking?, the device landscape in our SmartNIC and DPU Vendor Guide, and the architecture split in A Tale of Three Networks: AI Networking Infrastructure. Quarterly updates land in our AI infrastructure category.