Optical communications upcycle expectations raised again, with COHR offering the best positioning into the earnings window

Overview

The report combines upcoming C2Q26 results, recent industry discussions, and LightCounting market forecasts to assess the near-term earnings risk-reward for COHR, LITE, and FN, and updates the long-term opportunity for datacom optics, NPO/CPO, and telecom and data center interconnect markets. The core view is that AI clusters are driving upgrades in optical connection speeds, corporate earnings momentum is likely to improve in the second half of 2026, and 1.6T and next-generation packaged optics will support industry growth through 2030.

Core views

At the company level, COHR has the clearest path for revenue, margin, and product mix improvement, and is likely to deliver a cleaner earnings beat and guidance raise; although LITE faces debates around rising exposure to low-margin transceivers, China InP supply, and CPO progress, these risks are unlikely to materialize meaningfully in the near term, and cautious expectations instead create a positive earnings window; FN’s June quarter may deliver only a standard beat, and its appeal depends on whether September-quarter guidance can show acceleration in NVDA and AWS transceivers, high-performance computing, and OCS businesses. At the industry level, 1.6T is the main driver of upward revisions to datacom market forecasts, pluggable transceivers dominate incremental growth before 2028, while NPO/CPO dominate market expansion after 2028.

Analysis framework

The report uses a combined approach of company earnings catalyst analysis, tracking of product and customer project ramps, relative valuation comparison, and third-party market size forecasts. In the near term, it focuses on comparing revenue, gross margin, product mix, capacity, and quarterly guidance; over the long term, it breaks down the market opportunity by transmission speed, optical technology architecture, application scenario, and cloud service customers.

Methodology notes

• Earnings analysisEarnings beat and guidance raise framework

  Compare market expectations, current-quarter operating performance, and next-quarter management guidance

  By assessing volume ramp in transceivers, telecom and data center interconnect, EML, high-performance computing, and industrial businesses, the framework determines whether companies can beat earnings expectations and raise future guidance.

• Market size analysisBottom-up TAM forecast

  Break down the total addressable market by speed, technology architecture, application scenario, and customer

  Using LightCounting forecasts, the report breaks down the datacom market into 800G, 1.6T, 3.2T, as well as pluggable, LRO/LPO, NPO, and CPO, and compares market increments across different phases.

• Valuation analysisForward P/E relative comparison

  Compare the risk-reward of the three companies based on P/E multiples corresponding to CY28 earnings per share

  COHR, LITE, and FN trade at approximately 23x, 22x, and 23x CY28 expected P/E, respectively; the report combines upside to earnings to judge COHR and LITE as more attractively valued.

Asset mapping & comparison

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

• Coherent Corp (COHR)

  The report’s preferred earnings catalyst name, rated Overweight

  Strengths

  Transceiver revenue is expected to accelerate in the second half of 2026, InP capacity expansion supports margin improvement, and telecom and data center interconnect, OCS, CPO, and industrial businesses provide multiple drivers.

  Weaknesses

  Growth realization depends on multiple product and capacity projects ramping as planned.

  Comparison

  Compared with LITE and FN, the earnings outlook is the clearest; approximately 23x CY28 expected P/E is still considered not expensive.

  Risks

  InP expansion, transceiver and OCS/CPO volume ramp may fall short of expectations, or industrial demand recovery may be weaker than expected.

• Lumentum (LITE)

  An Overweight-rated name with a positive earnings window

  Strengths

  High-margin telecom, data center interconnect, and EML businesses may offset gross margin pressure from rapid Google transceiver volume ramp, while cautious market sentiment also lowers the expectations bar.

  Weaknesses

  Near-term product mix may be affected by rising exposure to lower-margin transceivers.

  Comparison

  Earnings certainty is slightly lower than COHR, but at approximately 22x CY28 expected P/E, it has the lowest multiple among the three companies and offers better risk-reward than FN.

  Risks

  Chinese competitors increasing InP supply, delays in scaled CPO projects, and transceiver volume ramp causing pressure on gross margin.

• Fabrinet (FN)

  A Neutral-rated name with positive long-term trends but weaker near-term risk-reward

  Strengths

  NVDA and AWS transceivers, AWS Trainium and high-performance computing, two new commercial transceiver projects, and OCS may drive upward revisions to forecasts for the second half of 2026 and 2027.

  Weaknesses

  The June quarter may deliver only a standard beat, evidence of EML supply improvement is limited, and earnings upside is lower than COHR and LITE.

  Comparison

  Approximately 23x CY28 expected P/E is close to COHR and LITE, but the earnings reaction depends more on September-quarter guidance, making its relative appeal the lowest.

  Risks

  September-quarter guidance fails to reflect acceleration in transceiver and high-performance computing, or EML supply, AWS projects, and OCS progress fall short of expectations.

• Nvidia (NVDA)

  An important driver of 1.6T datacom optics demand and a customer for FN-related projects

  Strengths

  Expected to account for more than 70% of the 1.6T market in 2026, driving its datacom market share to rise to 25% in the near term.

  Weaknesses

  Long-term share may decline as cloud service providers increase third-party procurement and diversify computing environments.

  Comparison

  The share rebound in 2026 is expected to be significantly stronger than the long-term trend, but market share may fall to 12% by 2030.

  Risks

  End-customer procurement diversification, 1.6T ramp falling short of expectations, and increased alternative supply solutions.

Key data

• Datacom optics market sizeApproximately $20 billion in 2025 and more than $70 billion in 2030Expected CAGR is close to 28%.
• 1.6T market sizeApproximately $40 billion in 2030Expected CAGR is approximately 120%, with related forecasts raised by an average of approximately 35% versus prior estimates.
• 3.2T market sizeApproximately $14 billion in 2030Expected to contribute relatively little before 2027, followed by rapid growth.
• NPO/CPO market sizeMore than $18 billion in 2030Expected to account for more than 25% of the datacom market by then, with the latest forecast raised by more than 50% on average versus prior estimates, mainly due to inclusion of the NPO market.
• Incremental market growth for pluggable transceivers from 2026 to 2028Approximately $14 billionDuring the same period, incremental growth for LRO/LPO, NPO, and CPO is each less than $3 billion.
• Incremental market growth for NPO and CPO from 2028 to 2030Approximately $9 billion and $6 billion, respectivelyExpected to become the main growth sources for the datacom market in this phase.
• Demand from the top five U.S. cloud service providersExpected CAGR of 29% through 2030Expected to contribute more than 60% of total demand in the datacom optics market, with Google and Meta leading growth.
• Nvidia datacom market share18% in 2025, 25% in 2026, and 12% in 2030The 2026 rebound comes from 1.6T volume ramp, with Nvidia expected to account for more than 70% of the 1.6T market that year.
• Telecom and data center interconnect market sizeApproximately $4 billion in 2025 and approximately $9 billion in 2030Expected CAGR is 18%, with the latest forecast raised by an average of 11%, mainly driven by 800G and above ZR/ZR+.
• FN’s AWS Trainium and high-performance computing businessExpected to reach $150 millionConsistent with prior guidance, and an important watch point for whether the September quarter can accelerate.

Impact & implications

In the near term, earnings catalysts and cautious expectations together increase the probability of positive share price reactions for COHR and LITE, with COHR offering the highest certainty in revenue growth, margin improvement, and product mix upgrade. In the medium term, 1.6T volume ramp will benefit InP, EML, transceivers, and related manufacturing supply chains. In the long term, optical connectivity will gradually expand from pluggable solutions to NPO/CPO, but trade-offs among architecture choice, serviceability, latency, and power consumption will determine supplier share. Demand growth from Google and Meta may change the customer mix, while Nvidia’s share rebound in 2026 is more likely to be a cyclical or phase-specific phenomenon.

Risks

• A rising revenue share of low-margin transceivers may pressure near-term gross margin.
• Chinese competitors increasing InP supply may intensify medium- to long-term pricing and share pressure.
• Scaled CPO, NPO, OCS, and new transceiver projects may face delays or ramps below expectations.
• FN’s share price reaction depends heavily on September-quarter guidance; if acceleration in transceivers and high-performance computing is not evident, valuation may come under pressure.
• NPO has longer copper interconnect paths and higher latency than CPO, which may limit its penetration in scale-out networks.
• 800G is expected to peak in 2028; if migration to 1.6T and 3.2T is slower than forecast, long-term market opportunity may be revised downward.
• Third-party market forecasts have been raised significantly due to the newly included NPO market, but uncertainty remains around actual adoption speed and architecture share.

What to watch

• COHR’s transceiver revenue growth, InP capacity utilization, margins, and improvement in industrial business orders.
• The impact of LITE’s Google transceiver ramp on gross margin, and whether telecom, data center interconnect, and EML businesses can provide offsets.
• LITE management commentary on scaled CPO, China InP competition, scale-out CPO, and OCS project progress.
• FN’s September-quarter revenue guidance and the degree of acceleration in NVDA and AWS transceivers, AWS Trainium, and high-performance computing businesses.
• Actual 1.6T market volume ramp in 2026, and whether Nvidia can reach expected market share of more than 70%.
• Commercial deployment of NPO/CPO after 2028, serviceability standards, and penetration in scale-out applications.
• Growth in optical procurement by Google and Meta, and whether Meta can become the largest NPO/CPO customer before 2030.
• Realization of upward forecast revisions for the telecom and data center interconnect market driven by 800G and above ZR/ZR+.