Limited short-term impact of GlassBridge on AI transceivers, but increased technology substitution risk for FAU

Overview

This report focuses on Corning's official launch of the GlassBridge glass optical interconnect component at the AI Data Center Optical Communication and Interconnect Technology Conference in Seoul on June 24, and discusses its potential implications for AI transceivers and FAU manufacturers as a Fiber-to-PIC connection platform. The report notes that GlassBridge is not an entirely new concept; related news had already emerged as early as September 2025, and it has been included in Corning's US$10 billion Photonics target mentioned at its Analyst Day.

Core views

The core view is that GlassBridge may become a potentially disruptive technology for existing FAU manufacturers in the development of CPO; however, its short-term impact on AI transceiver companies should be relatively limited, especially over the next 1-2 years. The reason is that GlassBridge can be used for both CPO and NPO, and if its application in NPO expands, it may partially offset the risks that the CPO path brings to traditional AI transceiver demand.

Analysis framework

The report uses technology-path comparison and industry-chain impact analysis: it first defines Corning GlassBridge and Fiber-to-PIC, then compares them with traditional FAU solutions in terms of high fiber count, assembly complexity, density, scalability, and system integration methods, and finally assesses the impact on FAU manufacturers, AI transceiver companies, and the CPO/NPO ecosystem respectively.

Methodology notes

• Technology path comparisonFiber-to-PIC and FAU substitution framework

  Treat GlassBridge as a Fiber-to-PIC connection platform and assess its substitutive or complementary role relative to traditional Fiber Array Unit solutions.

  The report believes that in extremely high fiber-count scenarios, traditional FAU faces rising assembly and scaling difficulty, while GlassBridge offers a higher-density and more scalable alternative through wafer-level passive alignment.

• Industry chain impactCPO/NPO application scenario breakdown

  Differentiate GlassBridge's applications in CPO and NPO to judge the net impact on different company types.

  FAU manufacturers face more direct technological disruption in the development of CPO; AI transceiver companies, by contrast, are judged to face limited short-term net impact because GlassBridge may also be used in NPO.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• AI transceiver companies

  Affected target

  Strengths

  Impact over the next 1-2 years is expected to be limited; GlassBridge can be used for NPO, which may bring some application expansion opportunities.

  Weaknesses

  If CPO is commercialized rapidly and changes the optical interconnect architecture, it may still affect the long-term demand structure for traditional transceivers.

  Comparison

  Compared with FAU manufacturers, AI transceiver companies face lower short-term risk.

  Risks

  Faster-than-expected CPO adoption, insufficient offsetting effect from NPO, changes in customer architecture choices.

• FAU makers

  Potentially substituted or disrupted target

  Strengths

  Existing FAU solutions remain an important current path for fiber-array connections.

  Weaknesses

  In extremely high fiber-count scenarios, traditional FAU assembly is complex and scaling becomes more difficult.

  Comparison

  Compared with AI transceiver companies, FAU manufacturers face more direct technology substitution risk from GlassBridge.

  Risks

  Accelerated commercialization of GlassBridge, CPO customers adopting new connection platforms, repricing of FAU-related option value.

• Corning GlassBridge

  New technology platform

  Strengths

  Supports higher density, better scalability, and detachable system integration, while reducing complexity through wafer-level passive alignment.

  Weaknesses

  The actual timetable for commercial deployment remains uncertain, and the report does not provide validation of mass production, cost curves, or the scale of customer adoption.

  Comparison

  Relative to traditional FAU solutions, GlassBridge offers a more direct Fiber-to-PIC connection path.

  Risks

  Commercialization progressing more slowly than expected, insufficient ecosystem compatibility, changes in CPO/NPO adoption paths.

Key data

• Report date2026-06-28The report body is dated June 28, 2026 02:31 PM GMT.
• Technology launch date2026-06-24Corning officially launched the GlassBridge glass optical interconnect component at the AI Data Center Optical Communication and Interconnect Technology Conference in Seoul.
• Short-term impact window1-2 yearsThe report expects very limited impact on AI transceiver companies over the next 1-2 years.
• Corning Photonics targetUS$10bnThe chart note mentions that GlassBridge has been included in Corning's US$10 billion Photonics target presented at Analyst Day.
• Morgan Stanley rating framework time horizon12-18 monthsThe disclosure section states that its stock ratings are generally based on risk-adjusted total return relative to industry coverage over the next 12-18 months.

Impact & implications

From an investment perspective, GlassBridge is more of a medium-term variable for AI optical interconnect manufacturing paths than an immediate short-term shock that changes the fundamentals of AI transceiver companies. For FAU manufacturers, if CPO adopts a higher-density, wafer-level, passively aligned Fiber-to-PIC solution, the technical barriers and demand elasticity of traditional FAU may be repriced; for AI transceiver companies, the near-term focus remains on whether broader NPO adoption can offset the potential substitution pressure brought by CPO.

Risks

• The actual commercialization timeline of GlassBridge remains uncertain, which may lead to valuation and share price volatility for related companies.
• FAU manufacturers face the risk of technological substitution or market-share pressure in the development of CPO.
• The judgment that AI transceiver companies will face limited short-term impact depends on NPO applications being able to partially offset CPO risk.
• The report does not provide specific company earnings estimates, target price changes, or quantified customer adoption data, so its investment conclusions still need to be validated with subsequent evidence.
• Morgan Stanley discloses that it has investment banking, shareholding, or service relationships with multiple covered companies, so potential conflicts of interest should be considered when using the research.

What to watch

• Subsequent customer validation, mass-production timeline, and cost curve for Corning GlassBridge.
• The pace of adoption of CPO and NPO architectures in AI data center optical interconnects.
• Changes in yield, assembly complexity, and pricing pressure for traditional FAU in high fiber-count scenarios.
• Whether AI transceiver companies disclose progress in adapting to Fiber-to-PIC, GlassBridge, or similar platforms.
• Valuation volatility and changes in order visibility for companies with high FAU-related option value.