SKN | Goldman Sachs Sees Optical Networking as AI’s Next Major Growth Opportunity

Key Points:

• Goldman Sachs projects the AI networking market will expand from approximately $15 billion to $154 billion by 2028 as optical technologies become essential for next-generation AI infrastructure.
• The firm expects co-packaged optics to account for roughly 59% of AI networking spending by 2028, driven by increasing bandwidth requirements.
• Goldman believes companies with significant exposure to optical networking, including Lumentum, could benefit from accelerating demand as hyperscale AI deployments expand.

Goldman Sachs believes the next major investment opportunity in artificial intelligence extends beyond processors to the networking infrastructure that connects increasingly powerful AI systems. As AI models grow larger and require more computing resources, the bank expects optical networking technologies to become a critical component of next-generation data centers.

The firm forecasts the addressable market for AI networking will increase from roughly $15 billion during the current technology cycle to approximately $154 billion by 2028, reflecting the rapid expansion of AI infrastructure spending.

Optical Technologies Become Increasingly Important

According to Goldman Sachs, traditional copper connections are reaching practical limitations as AI clusters require higher bandwidth, lower latency, and greater energy efficiency. Optical networking technologies, including pluggable transceivers and co-packaged optics, are expected to replace conventional electrical connections across many high-performance computing environments.

The transition is expected to accelerate as future AI platforms require significantly greater networking capacity to support communication among thousands of processors operating simultaneously.

Lumentum Positioned to Benefit From Industry Expansion

Goldman Sachs identifies Lumentum as one of the companies with the greatest direct exposure to the optical networking theme. Unlike diversified semiconductor manufacturers, Lumentum generates a substantial portion of its business from optical communication technologies, including lasers, photonic components, and optical engines used in high-speed data center networking.

As hyperscale cloud providers migrate toward faster 1.6-terabit and future 3.2-terabit optical connections, demand for these technologies could increase significantly if AI infrastructure investment continues at its current pace.

Co-Packaged Optics Represent the Next Phase of AI Infrastructure

The report projects that co-packaged optics could account for approximately $91 billion, or 59% of total AI networking spending, by 2028. This technology integrates optical components directly alongside computing processors, improving performance while reducing power consumption and transmission bottlenecks.

Although widespread commercial deployment remains subject to manufacturing, packaging, and customer qualification timelines, Goldman believes both existing pluggable optical solutions and future co-packaged architectures present meaningful long-term growth opportunities.

Execution Risks Remain

Despite the optimistic outlook, Goldman acknowledges that adoption timelines for co-packaged optics could extend beyond current projections. Manufacturing complexity, production yields, serviceability, and customer deployment schedules may slow commercialization.

The firm’s forecasts also assume continued robust investment by hyperscale cloud providers. Any moderation in AI infrastructure spending could reduce networking demand, although the long-term trend toward higher-performance data center connectivity is expected to remain intact.

Outlook

Goldman Sachs believes the next phase of artificial intelligence investment will increasingly focus on the infrastructure that enables large-scale AI systems to communicate efficiently. As demand shifts toward faster optical networking solutions, companies specializing in photonics and high-speed connectivity may become important beneficiaries alongside traditional semiconductor manufacturers. While adoption risks remain, the continued expansion of AI computing is expected to support long-term growth across the optical networking ecosystem.

For a confidential discussion regarding AI infrastructure investment trends, semiconductor supply chains, optical networking technologies, or institutional technology investment strategies, contact our senior advisory team.