FCC New Rules Take Effect: Optical Module Leaders Like Innolight Not Yet Listed
nashnova research
The FCC's new rule extends equipment review down to internal logic-bearing components, but none of the four major Chinese optical-module firms — including Zhongjichuangxin and Eoptolink — currently appear on the Covered List. The real test lies in next-generation product design wins.
What exactly did this rule change?
FCC final rule 26-50 was published in the Federal Register on September 11, 2026 and takes effect 30 days later.
The old framework only covered finished equipment: whoever built the box applied for authorization. Covered List firms locked out of whole-device authorization could retreat upstream, sell only the core digital module, and let someone else file — the so-called "component evasion loophole."
The new rule closes that loophole by defining "logic-bearing hardware components" — any device, module, or sub-assembly that uses digital technology, generates timing signals above 9,000 cycles per second, or uses RF energy for computing, storage, or data transmission.
This means → "sell it as parts" no longer bypasses the ban — but this is not a blanket prohibition on all Chinese components.
Are Chinese optical-module firms actually restricted?
A search of the current Covered List shows Zhongjichuangxin (中际旭创), Eoptolink (新易盛), O-Net Technologies (天孚通信), and Dongshan Precision (东山精密) are not named. The list currently covers Huawei, ZTE, Hikvision, Hytera, and others.
Xingzheng Securities' telecom team concluded: "The FCC shoe has dropped; optical-communications sector fears should ease significantly."
A product triggers the ban only if it meets all four conditions simultaneously: the component is logic-bearing + it is made by a Covered List entity + that entity's finished device would also be banned + the application falls under the new rule's scope. Miss any one, and the ban does not apply.
In plain terms = the words "Chinese optical module" alone do not trigger the ban — what matters is who made it, what it does, and whether the maker is on the list.
Inside a module, which parts face higher risk?
A high-speed optical module contains digital logic, analog circuitry, and optical devices side by side. Regulatory risk is clearly tiered.
Higher risk: DSP — digital signal processor, MCU — microcontroller, CDR/Retimer — clock-recovery and signal-relay chips. These carry digital logic or data-processing capability, closest to the FCC's definition. An integrated optical module containing such functions may itself be deemed logic-bearing.
Lower risk: CW lasers, standalone EML — electro-absorption modulated lasers, lenses, FAU — fiber array units, fiber connectors. If these only generate, modulate, or transmit optical signals without digital logic, the likelihood of falling directly under the ban is currently low.
This means → within the same module, the chip side and the optics side face entirely different compliance pressure.
How much does this affect existing orders and shipping products?
The new rule applies prospectively: already-authorized equipment will not lose its status retroactively. Pending applications can, in principle, proceed — but adding or swapping a logic-bearing hardware component afterward may trigger treatment as a new application.
For 800G and 1.6T optical modules, near-term shipments are driven more by existing customer orders and AI data-center build schedules. With leading Chinese suppliers not on the list and already-qualified platforms not subject to retroactive removal, the rule is unlikely to translate directly into current-quarter revenue decline.
In plain terms = the installed base faces limited immediate disruption — there will be no sudden order cancellations.
Where is the real risk window?
3.2T, NPO (near-package optics), and CPO (co-packaged optics) are still in early design or qualification. Their bills of materials have not yet been locked down.
System vendors can more easily write sourcing restrictions into design specs at this stage, favoring suppliers whose compliance path is clearest among those with similar performance.
This means → the new rule's impact on next-generation products will most likely show up first in design wins, sampling, and certification cycles — not as a sudden drop in quarterly earnings.
How do compliance costs ripple through? Whose bargaining power grows?
After the rule takes effect, system applicants must prove where their key logic hardware comes from. The first change module makers will feel is likely not a letter from the regulator but customers adding supplier declarations, affiliated-entity checks, BOM audits, and component-change notification requirements.
Even if the law does not ban a particular supplier, customers may pre-emptively tighten their approved-vendor lists to avoid future re-qualification costs. This reflects a shift: compliance is moving from "get a certificate" to "continuously prove your supply chain."
Larger suppliers with more R&D and compliance resources adapt more easily. Smaller firms, even if unnamed, may lose projects simply because they cannot provide full traceability documentation.
The FCC is separately discussing hardware bills of materials (HBOM), software bills of materials (SBOM), white-box equipment, and authorization durations — none of which are part of this final rule yet. But the policy direction is clear: equipment authorization is evolving from a one-time certificate into an ongoing, auditable supply-chain file.
市场有风险，内容仅供研究参考，不构成投资建议。