OFC 2026 Preview: How Silicon Photonics and CPO Will Redefine the Next Generation AI Interconnection System

Optical Interconnect Is Becoming the Key Bottleneck and Opportunity in AI Infrastructure.

As OFC 2026 approaches this month in Los Angeles, the global optical communications industry will showcase the latest advancements in next-generation connectivity technologies. Driven by the strong demand for bandwidth in AI data centers, 800G optical modules are moving from pilot projects to mainstream adoption, with 1.6T products entering mass production ramp-up. Optical interconnect technology is undergoing a structural transformation.

According to SemiVision Research, the rapid expansion of generative AI model sizes is shifting the core bottleneck in data centers from transistor performance to interconnect bandwidth and latency. Nvidia CEO Jensen Huang’s “Huang’s Law” states that computing power can be continuously improved through process advancements and 3D packaging, but the I/O rate between chips lags behind, creating an “I/O wall.” This structural contradiction is driving optical devices to migrate toward compute and switching chips to overcome power consumption, loss, and transmission distance constraints.

The OFC technical conference will be held from March 15 to 19, with the exhibition opening at the Los Angeles Convention Center from March 17 to 19. Key topics include: 1.6T and 3.2T optical module technology pathways, co-packaged optics (CPO) and new optical I/O architectures, silicon photonics heterogeneous integration, as well as the competition between pluggable solutions and CPO routes.

Copper Cables Reach Physical Limits, Optical Interconnects Shift to Chip Side

The exponential growth of AI clusters is pushing optical interconnect technology to the core of infrastructure architecture. Market research firm SemiVision Research notes that traditional data centers typically use copper cables for short-distance rack-to-rack connections, while pluggable optical modules handle inter-rack expansion links. However, as SerDes speeds increase to 200 Gb/s per channel, the physical limitations of copper cables become a critical bottleneck.

According to Business Wire, at these speeds, traditional passive copper cables can no longer reliably span a single server rack, and even within racks, transmission distances are limited. These physical constraints are driving optical devices to continue migrating toward compute and switching chips to achieve a new balance among power consumption, loss, and transmission distance.

800G Going Mainstream, 1.6T Mass Production Nears

From product evolution, the optical module market is showing a clear upgrade rhythm: 800G products entered a strong growth phase in 2025 and are accelerating into mainstream markets; 1.6T products have been in mass production ramp-up since the second half of this year.

Specifically, Accelink has publicly demonstrated and shipped samples of 1.6T OSFP224 DR8 transceivers designed for AI data center scenarios. SemiVision Research points out that as AI cluster sizes rapidly expand to hundreds of thousands of GPUs, optical interconnects have become one of the most critical bottlenecks in AI infrastructure stacks and a highly valuable investment track.

CPO and Pluggable Solutions Competition Becomes Focus

The competition between co-packaged optics (CPO) and pluggable solutions will be one of the core topics at OFC 2026. CPO integrates optical engines and switching chips on the same substrate, significantly reducing power consumption and signal loss, and is viewed as the long-term evolution direction for large-scale data centers; pluggable solutions, with their flexibility and maintainability, still dominate the current stage.

SemiVision Research reveals that Nvidia, Broadcom, and Marvell will present on manufacturability issues at this OFC, which directly relate to the timeline for CPO moving from technical validation to large-scale commercial deployment. Meanwhile, the evolution path from pluggable to linear pluggable optics (LPO), and then to CPO, along with the resulting supply chain restructuring, will also be focal points of the exhibition.

Silicon Photonics and Laser Technologies as Key Supply Chain Variables

At the foundational technology level, silicon photonics heterogeneous integration—including platforms combining silicon photonics, thin-film lithium niobate (TFLN), and III-V materials—is a key focus for OFC 2026. SemiVision Research notes that the supply-demand bottleneck of laser technology has become a critical factor constraining the expansion of optical interconnects.

Additionally, the competition between VCSEL and MicroLED technologies in ultra-short-distance optical interconnects, as well as optical interconnect solutions replacing copper cables inside AI systems, will attract widespread attention. Notably, optical I/O (OIO), as a native optical link technology in packaging, is expected to support decoupled deployment of AI systems and is a promising emerging direction to watch.

Outlook: Optical Interconnect Enters a Strategic Competition Window

As OFC 2026 approaches, it coincides with a critical point where the demand for optical interconnects in AI data centers is accelerating. From the product iterations of 800G to 1.6T, from architecture evolutions between pluggable and CPO, to the shift from copper to optical short-distance solutions, multiple technological trajectories are advancing simultaneously, building the foundation for this round of optical interconnect revolution.

SemiVision Research believes that, as AI cluster sizes continue to expand into the hundreds of thousands of GPUs, optical interconnects have surpassed traditional supporting infrastructure roles and are becoming a core variable determining the upper performance limit of AI infrastructure. The technological consensus and industry trends emerging from this OFC will have a profound impact on the future landscape of optical communications in the coming years.

Risk Warning and Disclaimer

Market risks exist; investments should be cautious. This article does not constitute personal investment advice and does not consider individual users’ specific investment goals, financial situations, or needs. Users should consider whether any opinions, viewpoints, or conclusions herein are suitable for their particular circumstances. Invest at your own risk.