AI data centers hit a new bottleneck: Optical interconnects

[Abstract Flow of Digital Data Streams in Blue Neon Light]
asbe

INVESTORS SHOULD KNOW: The AI chip supply chain is facing a critical bottleneck at the manufacturing and testing layer, where optical interconnect scaling and silicon photonics capacity are emerging as key constraints on AI data center buildout.

BACKGROUND

As AI data centers scale to handle larger workloads, the physical infrastructure connecting chips inside those facilities has become a limiting factor. Optical interconnects, which use light rather than electrical signals to move data, are increasingly seen as the solution to bandwidth and power constraints at scale.

The shift toward co-packaged optics and silicon photonics is not just a design challenge. It is a manufacturing and testing challenge. Equipment suppliers, test solution providers, and contract manufacturers all play roles in determining how quickly optical interconnect technology can reach high-volume production.

STOCKS TO WATCH

Key publicly traded companies in this supply chain include ASML Holding N.V. (ASML [https://seekingalpha.com/symbol/ASML]) and Lam Research Corporation (LRCX [https://seekingalpha.com/symbol/LRCX]), both core semiconductor equipment suppliers whose tools are used in advanced chip fabrication. ASM International NV (ASMIY [https://seekingalpha.com/symbol/ASMIY]) and Tokyo Electron Limited (TOELY [https://seekingalpha.com/symbol/TOELY]) are also major equipment players in the space. On the components and connectivity side, Credo Technology (CRDO [https://seekingalpha.com/symbol/CRDO]) and MaxLinear (MXL [https://seekingalpha.com/symbol/MXL]) are among the companies reporting direct revenue exposure to optical data center ramps. Fabrinet (FN [https://seekingalpha.com/symbol/FN]) serves as a contract manufacturer for optical and datacom products.

Additional companies with exposure to the AI chip supply chain theme include Nova Ltd. (NVMI [https://seekingalpha.com/symbol/NVMI]), MKS Instruments (MKSI [https://seekingalpha.com/symbol/MKSI]), QuickLogic Corporation (QUIK [https://seekingalpha.com/symbol/QUIK]), Atomera Incorporated (ATOM [https://seekingalpha.com/symbol/ATOM]), ChipMOS Technologies (IMOS [https://seekingalpha.com/symbol/IMOS]), and X-FAB Silicon Foundries SE (XFABF [https://seekingalpha.com/symbol/XFABF]), spanning process control, materials, and foundry services.

KEY TAKEAWAYS

* Optical interconnect scaling is increasingly a manufacturing and testing bottleneck, not just a design problem — creating demand for specialized equipment and test infrastructure across the AI chip supply chain.
* Several companies in the space are reporting accelerating revenue tied to optical data center ramps, with supply constraints and component availability cited as near-term execution risks.
* New partnerships and equipment order announcements signal that the photonics manufacturing ecosystem is actively expanding capacity to meet AI data center demand.

NEWS ROUNDUP

Recent announcements and earnings commentary from across the AI chip supply chain illustrate both the scale of the opportunity and the real-world constraints companies are navigating. Revenue targets are rising, but supply tightness and component availability remain active concerns for management teams and analysts alike.

* CREDO TECHNOLOGY (CRDO [https://seekingalpha.com/symbol/CRDO]) — Management shifted its fiscal 2027 growth outlook to more than 80% year-over-year revenue growth [https://seekingalpha.com/news/4599516-credo-signals-80-percent-fiscal-2027-revenue-growth-with-600m-expected-from-optical-portfolio], with over $600M expected from its optical portfolio. Management also flagged supply chain tightness and customer concentration as ongoing considerations.
* MAXLINEAR (MXL [https://seekingalpha.com/symbol/MXL]) — Management raised its 2026 optical data center revenue target to $150M–$170M [https://seekingalpha.com/news/4579167-maxlinear-outlines-160m-170m-q2-revenue-outlook-as-it-lifts-2026-optical-data-center-target], citing improved visibility into 400G and 800G ramps. Wafer costs and packaging constraints were cited as factors affecting margin sensitivity.
* FABRINET (FN [https://seekingalpha.com/symbol/FN]) — Management described strong DCI growth trends alongside near-term supply constraints impacting Datacom shipments [https://seekingalpha.com/news/4585330-fabrinet-forecasts-1_25b-1_29b-q4-revenue-as-datacom-supply-constraints-persist]. Analysts pressed management on execution risk tied to component availability.
* ADVANTEST (ATEYY [https://seekingalpha.com/symbol/ATEYY]) — The company announced a partnership with OpenLight to develop silicon photonics test solutions [https://seekingalpha.com/pr/20559785-advantest-partners-with-openlight-to-develop-silicon-photonics-test-solutions-for-high-volume] aimed at scalable, high-volume manufacturing for electro-optical devices used in AI and HPC data centers. High-volume testing is a key gating factor for broader silicon photonics adoption.
* CLASSONE TECHNOLOGY — The company reported record Solstice S8 orders from Applied Optoelectronics [https://seekingalpha.com/pr/20560257-classone-technology-secures-record-solstice-s8-orders-from-applied-optoelectronics], tied to the expansion of 6-inch InP production for optical interconnect applications. The announcement frames the shift toward co-packaged optics as a major architectural transition for AI data centers.

AI CHIP SUPPLY CHAIN IN FOCUS

The screener below identifies publicly traded companies aligned with the AI chip supply chain theme, ranked by relevance to the theme, with the most theme-relevant companies appearing first.

{"type":"screeners","props":{"type":"stocks","name":"AI Chip Supply Chain","filter":{"ds_screener_filter":{"in":["ai_bottlenecks_chip_supply_chain"]}},"sort":"ds_screener_filter"},"data":[{},{},{},{},{},{},{},{},{"slug":"nvmi","ds_screener_filter":"AI: Chip Supply Chain","company_name":"Nova Ltd.","has_virtual_analysis":true},{"slug":"xfabf","ds_screener_filter":"AI: Chip Supply Chain","company_name":"X-FAB Silicon Foundries SE","has_virtual_analysis":false}],"show_in_screener_page":true}

The screener results span semiconductor equipment, process control, foundry services, and specialty chip design — covering ASM International NV (ASMIY [https://seekingalpha.com/symbol/ASMIY]), Tokyo Electron Limited (TOELY [https://seekingalpha.com/symbol/TOELY]), QuickLogic Corporation (QUIK [https://seekingalpha.com/symbol/QUIK]), Atomera Incorporated (ATOM [https://seekingalpha.com/symbol/ATOM]), MKS Instruments (MKSI [https://seekingalpha.com/symbol/MKSI]), Lam Research Corporation (LRCX [https://seekingalpha.com/symbol/LRCX]), ASML Holding N.V. (ASML [https://seekingalpha.com/symbol/ASML]), ChipMOS Technologies (IMOS [https://seekingalpha.com/symbol/IMOS]), Nova Ltd. (NVMI [https://seekingalpha.com/symbol/NVMI]), and X-FAB Silicon Foundries SE (XFABF [https://seekingalpha.com/symbol/XFABF]). The list reflects the breadth of the supply chain, from lithography and etch equipment at the top to foundry and specialty process players further down.

MORE ON

* PNC Financial Vs. U.S. Bancorp: The Total-Return Case Has Flipped [https://seekingalpha.com/article/4917077-pnc-financial-vs-us-bancorp-the-total-return-case-has-flipped]
* SM Energy Company (SM) Presents at J.P. Morgan Energy, Power & Renewables Conference 2026 Transcript [https://seekingalpha.com/article/4917064-sm-energy-company-sm-presents-at-j-p-morgan-energy-power-and-renewables-conference-2026]
* VICI Properties: Investors Are Misunderstanding A Historic Opportunity [https://seekingalpha.com/article/4917076-vici-properties-investors-misunderstanding-historic-opportunity]
* APLE, DHC among top 10 mid-cap U.S. REIT stocks by momentum grade [https://seekingalpha.com/news/4606230-aple-dhc-among-top-10-mid-cap-us-reit-stocks-by-momentum-grade]
* Top small-cap U.S. REIT stocks by momentum grades: Check who topped the chart [https://seekingalpha.com/news/4606205-top-small-cap-us-reit-stocks-by-momentum-grades-check-who-topped-the-chart]