800G Optical Transceivers: Why AI Clusters Drove a USD 23.8 Billion Optics Year
Updated: 4 days ago
Executive summary. The optical transceiver used to be a quiet component in the data centre: a small pluggable module that turned electrical signals into light so servers and switches could talk. Artificial intelligence changed its status. Training and running large AI models links thousands of accelerators that must exchange data constantly, and every link needs optics. In 2025 the transceiver became one of the most important constraints on how fast AI capacity could be built. Strategic Revenue Insights estimates the global 800G optical transceiver market at USD 5.80 billion in 2025, rising to USD 30.36 billion by 2035 at an 18.0 percent compound annual growth rate. AI cluster build-out, hyperscale switch upgrades, the scaling of silicon photonics and pressure for power efficiency are driving it. This article sets out five lessons from the module makers, chip suppliers and hyperscale buyers moving first, with takeaways for strategy, technology, supply chain, product and revenue.
The structure of the market explains where value is concentrating. Strategic Revenue Insights' 800G Optical Transceiver Market research finds that QSFP-DD is the largest form factor at 55.00 percent of 2025 value while OSFP is the fastest growing, that short reach is both the largest and the fastest growing reach segment, that EML lasers are the largest technology while silicon photonics is the fastest growing, and that data centres are the largest application while AI and high-performance computing clusters are the fastest growing. Asia Pacific is both the largest region, at 52.07 percent of 2025 value, and the fastest growing. The research names laser and EML supply limits and the shift to 1.6T as restraints, and lists Innolight, Eoptolink, Coherent, Cisco Systems, Broadcom, Lumentum, Marvell, Applied Optoelectronics, Accelink and Hisense Broadband among the leading companies.
Industry data shows the scale of the AI optics boom. LightCounting estimates that sales of optical transceivers and related products reached USD 23.8 billion in 2025, up 55 percent from 2024, and that optics used in AI clusters generated about USD 16.5 billion in 2025, with a projection of around USD 26 billion in 2026. LightCounting also ranked Innolight first in global transceiver shipments for 2025, with about 23.4 percent share. The market has moved from steady telecom-style growth to a capacity race paced by AI infrastructure spending.
Why the 800G Optical Transceiver Market Is Entering a New Phase
Three structural forces are converging.
AI clusters multiply optical links. An AI training cluster connects thousands of accelerators through several layers of switches, and every connection needs a transceiver at each end. SRI's research names AI cluster build-out as the largest driver of growth and identifies AI and HPC clusters as the fastest growing application. More accelerators per cluster means many more optical modules per server than in a traditional data centre.
Switches have moved to faster ports. Hyperscale operators are upgrading their networks to 51.2 terabit switches with 800G ports. SRI's research names hyperscale switch upgrades as a major driver. Each switch generation resets demand for a new class of transceiver.
Power has become a design limit. Optics consume a growing share of data centre power as speeds rise. SRI's research names power efficiency pressure as a driver and lists linear pluggable and low-power optics as an opportunity. Watts per gigabit now matters as much as speed.
These forces explain SRI's forecast of rapid growth to 2035, with value concentrating in silicon photonics, AI clusters and Asia Pacific manufacturing.
Five Lessons From the Companies Moving First
The module makers, chip suppliers and buyers turning these forces into advantage share a recognisable pattern. We have distilled it into five lessons, each followed by a takeaway for a specific leadership function.
Lesson 1: Volume and Qualification Win Hyperscalers
The buyers of 800G transceivers are concentrated: a handful of hyperscale cloud and AI companies account for most demand. SRI's research names a concentrated customer base with strong bargaining power as a threat. Winning these customers depends on passing long qualification processes and then delivering millions of modules on schedule.
The companies moving first invest in manufacturing scale, automated test and close engineering collaboration with hyperscale customers during design. Being qualified early on a new switch or accelerator platform locks in volume for that generation. Innolight's leading share in 2025 reflects that combination of scale and early qualification.
The infrastructure behind AI agents and enterprise AI, discussed in why the scaling gap, not the model, decides who wins in AI agents, ultimately depends on this physical layer being delivered on time.
"In AI optics, the product that wins is the one already qualified and shipping by the million when the next cluster is built."
Takeaway on strategy. Invest in manufacturing scale and automated test, engage hyperscale customers early in platform design, and plan capacity around their build schedules. Treat qualification as the main sales process.
Lesson 2: Silicon Photonics Is the Scale Technology
SRI's research identifies EML lasers as the largest technology today and silicon photonics as the fastest growing, and names silicon photonics scale as a driver. Silicon photonics builds optical components on silicon wafers using semiconductor processes, which can lower cost, improve integration and ease laser supply constraints.
The companies moving first combine silicon photonics design, access to foundry capacity and laser supply partnerships. SRI's research names laser and EML supply limits as the largest restraint, and silicon photonics reduces the number of high-end lasers needed per module. Vertical integration across lasers, photonic chips and digital signal processors is becoming a competitive advantage.
Takeaway on technology. Build silicon photonics capability with secure foundry access, reduce dependence on scarce EML lasers, and integrate lasers, photonic chips and signal processing where it improves cost and supply resilience.
Lesson 3: Power per Bit Is the New Battleground
As data rates rise, optics consume more of the data centre's power budget. SRI's research names power efficiency as a driver and lists linear pluggable and low-power optics and co-packaged optics pilots with switch vendors as opportunities. Linear pluggable optics remove or simplify the digital signal processor to save power, while co-packaged optics place the optics next to the switch chip.
The companies moving first offer a range: conventional modules for compatibility, linear pluggable options for power savings and co-packaged optics pilots for the next architecture. Power limits link optics to the wider energy challenge of AI data centres described in how intelligent infrastructure is reshaping the AI in energy growth cycle.
"Every watt saved in the optics is a watt available for compute. In AI data centres, power per bit has become a product specification buyers negotiate on."
Takeaway on product. Offer low-power options, including linear pluggable optics, alongside standard modules, invest in co-packaged optics with switch partners and market watts per gigabit as a primary specification.
Lesson 4: Plan for Price Erosion and the Shift to 1.6T
SRI's research names rapid price erosion as volumes rise and faster migration to 1.6T modules as threats, and the shift to 1.6T as a restraint on 800G growth. Each generation follows a familiar pattern: high prices at launch, rapid volume growth, then price declines as competitors qualify and the next generation arrives.
The companies moving first manage the transition deliberately: cost reduction roadmaps for 800G, early development of 1.6T, and second-source supply programmes for hyperscale buyers, which SRI's research lists as an opportunity. They also extend 800G into new uses such as 800G ZR coherent modules for data centre interconnect, another opportunity in SRI's research.
Takeaway on supply chain. Build cost-down roadmaps for each generation, launch 1.6T early, offer second-source supply to hyperscale buyers and extend 800G into data centre interconnect and enterprise uses as AI clusters move to 1.6T.
Lesson 5: Asia Pacific Manufactures, Trade Policy Decides Access
SRI's research identifies Asia Pacific as both the largest and the fastest growing region. Chinese module makers such as Innolight and Eoptolink lead global shipments, and manufacturing is concentrated in China and Southeast Asia. Demand, meanwhile, is concentrated among US hyperscale buyers.
That creates exposure to trade policy. SRI's research names export controls and trade tariffs on optical parts as a threat. The companies moving first diversify manufacturing across countries, particularly into Southeast Asia, build supply chains that comply with export rules and maintain capacity close to key customers.
The capabilities that decide who captures this demand are specific:
Manufacturing scale and early qualification with hyperscale customers.
Silicon photonics design and secure laser and foundry supply.
Low-power and co-packaged optics roadmaps.
Cost-down plans and early 1.6T products.
Diversified, trade-compliant manufacturing footprints.
The enterprise connectivity layer that sits beyond the data centre is discussed in why private 5G networks are becoming the factory's nervous system. Companies evaluating these competitive dynamics in greater detail can use SRI's 800G Optical Transceiver Market analysis to assess form factor, reach, technology and application segments and the positioning of the leading suppliers.
Takeaway on market development. Diversify manufacturing across Asia Pacific, especially Southeast Asia, build trade-compliant supply chains and keep capacity aligned with the build plans of North American hyperscale customers.
Where the Opportunity Is Concentrating
Value is moving toward AI clusters, silicon photonics and low-power designs.
By form factor, QSFP-DD is the largest segment while OSFP is the fastest growing, driven by AI cluster designs.
By reach, short reach is both the largest and the fastest growing segment, reflecting dense connections inside AI clusters.
By technology, EML lasers are the largest segment while silicon photonics is the fastest growing.
By application, data centres are the largest segment while AI and HPC clusters are the fastest growing, and by geography Asia Pacific is both the largest and fastest growing region.
At the Horizon: What to Watch Over the Next Three Years
The 1.6T ramp. How quickly AI clusters move to 1.6T will determine the length of the 800G volume cycle.
Co-packaged optics adoption. Production deployments of co-packaged optics would change who captures value between module makers and switch chip vendors.
Trade policy. Tariffs and export controls on optical components will shape where modules are made and who can supply which customers.
"The next phase of AI infrastructure will not be limited by how many accelerators can be built. It will be limited by how fast, efficiently and reliably they can be connected."
The Strategic Revenue Opportunity
For module makers, the task is to scale manufacturing, adopt silicon photonics and plan each generation's transition. For chip and laser suppliers, the opportunity is in photonic integration and low-power designs. For hyperscale buyers, the priority is secure, diversified supply. For investors, the signal is qualification wins, capacity expansion and 1.6T readiness.
Market intelligence turns these questions into decisions. Segment-level sizing shows where form factor, technology and application demand is concentrating, competitive analysis shows how suppliers are positioned, and regional analysis shows where manufacturing and trade policy will shape supply. This is the path from market signals to strategic insight to revenue impact.
Takeaway on revenue. Treat the 800G optical transceiver market as four revenue pools: high-volume AI cluster modules sold through qualification, data centre interconnect coherent modules sold on reach, low-power and co-packaged optics sold on power savings, and second-source supply sold to hyperscalers on resilience. Each needs its own pricing model.
The Next Leg: Preparing for the Journey Ahead
The defining shift in optical transceivers is that they have moved from commodity components to a constraint on AI capacity. That changes how they are designed, bought and supplied. The companies moving first are not the ones with the most product variants; they are the ones winning qualification at scale, adopting silicon photonics, cutting power per bit, managing generational transitions and diversifying manufacturing.
The next three years will decide which suppliers become the long-term optical partners of the AI build-out. Decision-makers should invest in scale and photonic integration, prepare for 1.6T and track the ramp of new generations, co-packaged optics and trade policy. Companies evaluating growth opportunities in this market can use Strategic Revenue Insights' 800G Optical Transceiver Market report to assess market size, competitive dynamics, regional opportunities and strategic growth priorities.
Sources: Strategic Revenue Insights, 800g Optical Transceiver Market report (2026); LightCounting quarterly market update (March 2026).
Frequently Asked Questions
How big is the 800G optical transceiver market?
Strategic Revenue Insights estimates the global 800G optical transceiver market at USD 5.80 billion in 2025, rising to USD 30.36 billion by 2035 at an 18.0 percent compound annual growth rate. Asia Pacific is both the largest region, at 52.07 percent of 2025 value, and the fastest growing.
What is driving demand for 800G optical transceivers?
The build-out of AI clusters, hyperscale switch upgrades, the scaling of silicon photonics and pressure for power efficiency are the main drivers. LightCounting estimates optical transceiver sales reached USD 23.8 billion in 2025, up 55 percent from 2024.
Which 800G transceiver segments lead the market?
QSFP-DD is the largest form factor, short reach is the largest reach segment, EML lasers are the largest technology and data centres are the largest application. OSFP, silicon photonics and AI and HPC clusters are the fastest growing segments, according to Strategic Revenue Insights.
What are the main risks in the 800G optical transceiver market?
Laser and EML supply limits, rapid price erosion as volumes rise, a concentrated customer base, faster migration to 1.6T modules and export controls and tariffs on optical parts are the main risks identified in Strategic Revenue Insights' research.
About the authors: This article was prepared by the Strategic Revenue Insights research team, which combines market intelligence, competitive analysis and strategic advisory for growth-stage, mid-market and enterprise organisations across the ICT, semiconductor and electronics sector. Sector analyst: Lukas Schmidt, Technology and Infrastructure Analyst. Our market estimates combine primary research (interviews, surveys and expert panels) with secondary data (industry reports, databases and public disclosures); read more about how Strategic Revenue Insights works. To discuss this market with our analysts, contact our team.
Comments