Goldman Sachs Says That AI Infra Megatrend Has 9x Growth

Hook: The Next Big AI Infrastructure Megatrend Could Be In The Wires

Artificial intelligence is no longer just a lab curiosity; it is reshaping how data centers are built and upgraded. Training giant models, running real-time inference, and delivering AI-driven services require a steady stream of data to move, store, and process at scale. That means the backbone—optical networking, high-speed memory, and related data-center components—must keep pace with explosive demand. When memory and bandwidth bottlenecks ease, AI hardware like accelerators can unlock their full potential, delivering faster results and lower costs per inference.

Why AI Infrastructure Could Be The Next Megatrend

In recent years, the AI boom has created an unusual tug-of-war between demand for memory and the ability to move data at the speed AI requires. Memory chips—especially high-bandwidth types—have faced supply constraints that tighten up data-center expansion cycles. While memory remains critical, a broader trend is unfolding: the need for faster, more efficient data transport across networks. That is the sweet spot for optical networking, which uses light to shuttle massive data streams across racks, switches, and long-haul links.

From a strategic standpoint, the total addressable market for optical networking components is expanding rapidly. Industry analysts point to a multi-year runway as hyperscalers, edge deployments, and telecom operators push capex to keep pace with AI workloads. A growing chorus of researchers and bankers argue that optical networking could be the next big megatrend in the AI infrastructure arena. Critics warn about cyclical volatility, but the trend lines look compelling for players tied to data movement and network upgrades.

Goldman Sachs Says That Optical Networking Could Multiply TAM By 9x

As part of a bullish assessment on AI infrastructure, goldman sachs says that the growing demand for high-speed data transport could propel the optical networking segment into a $154 billion market. The logic hinges on a simple, yet powerful idea: as models get larger, the speed at which data can be moved between CPU/GPU accelerators and memory must scale just as fast. Without this, the performance gains from AI accelerators aren’t fully realized. The note adds that the TAM expansion could unfold over a relatively short horizon, helped by greenfield deployments and upgrades in data centers and networks.

Two key drivers underlie this forecast: first, the hyperscale operators are continuing to upgrade networks to 400G and beyond; second, carriers are accelerating backbones to accommodate 800G and new modulation formats that squeeze more data through the same fiber. When you combine these drivers with ongoing supply chain improvements and the broader AI demand cycle, the case for optical networking as a megatrend grows stronger. In other words, the thesis isn’t just about one component; it’s about a network upgrade cycle that could last several years.

Three Stocks You Can Buy to Capitalize on the $154 Billion Opportunity

Below are three names that sit at the heart of optical networking and AI infrastructure. They offer different angles on the same trend: a combination of product leadership, healthy balance sheets, and exposure to data-center and network upgrades. This isn’t financial advice—it’s a framework to guide your own due-diligence and risk evaluation.

Ciena Corporation (CIEN)

• Ciena is a recognized leader in optical transport gear, with a broad portfolio spanning metro, regional, and long-haul networks. Its coherent optics and software-driven network solutions are central to the AI-era growth in data movement.
• growth catalysts: 400G/600G deployments accelerate as cloud operators refresh to support larger AI models. The company’s software automation and analytics platform helps operators optimize capacity and energy use, which is attractive to hyperscalers watching unit economics.
• risks: Cyclical capex cycles, competition from newer entrants, and supply-chain volatility could pressure margins in the near term.
• how to think about value: Look for steady free-cash-flow generation, a manageable debt load, and backlog that translates into revenue visibility for 2025–2027. A multiple expansion could come if AI-driven data movement remains persistent and orders beat expectations.

Pro Tip: Track Ciena’s bookings and net backlog trend quarter-to-quarter. A steady or rising backlog often signals durable demand from large network rebuilds, which bodes well for revenue predictability.

Lumentum Holdings (LITE)

• why it helps: LUMENTUM is a core supplier of optical components—lasers and transceivers—used in data-center and telecom networks. Its products enable high-speed data transfer, making it a direct beneficiary of AI-driven fiber upgrades.
• growth catalysts: The rise of 400G and 800G transceivers, plus stronger demand from hyperscalers, support higher ASPs and more recurring revenue through component upgrades and warranty/maintenance streams.
• risks: Dependence on a handful of customers and commodity input costs can impact margins; also, technological shifts could alter supplier mix.
• how to think about value: Evaluate gross margin resilience and free cash flow. LITE’s exposure to data-center refresh cycles makes it sensitive to capex sentiment, but its diversified product lines can cushion volatility.

Pro Tip: Compare LUMENTUM’s newer transceiver platforms against the incumbent players. If you see improving yields and customer win rates, that’s a sign the AI network upgrade cycle is broadening.

Nokia Corporation (NOK)

• why it helps: Nokia supplies telecom-grade optical gear and software that power large-scale networks, including 5G backbones and edge solutions. As operators invest in capacity for AI workloads, Nokia gains from improved fiber access, routing, and disaggregation trends.
• growth catalysts: 400G/800G upgrades, the shift to software-defined networks, and O-RAN initiatives create multi-year demand tails beyond traditional hardware refreshes.
• risks: The competitive landscape includes larger players with entrenched positions, foreign-currency exposure, and the risk that telecom capex fans out differently across regions.
• how to think about value: Look for robust dividend support, free cash flow yields, and evidence of customers converting from legacy to AI-friendly upgrades.

Pro Tip: Nokia’s performance can reflect global capex sentiment. If you see telecoms accelerating 5G and fiber refresh cycles, NOK often benefits from both service-provider budgets and enterprise connectivity upgrades.

What to Watch: Catalysts, Risk, and Timing

The thesis behind investing in optical networking tied to AI is straightforward: more data movement means more demand for high-speed, reliable networks. That said, timing matters. Here are practical signals to monitor as you consider positions in CIEN, LITE, and NOK.

• Capex cycles: A ramp in hyperscaler and telecom budgets typically precedes a rebound in orders for optical gear. Watch quarterly capex commentary and backlog progression.
• Technology refresh cadence: Upgrades to 400G/800G transceivers and coherent optics tend to cluster, creating bursts of demand that can lift revenue visibility for several quarters.
• Supply-chain health: Chip and semiconductor supplier lead times, pricing trends, and component availability can affect gross margins and product mix.
• Regulatory and geopolitical factors: Trade dynamics and export controls can influence supply chains for specialized optics, so staying aware of policy shifts helps manage risk.

Pro Tip: Build a watchlist with three to five stock-price catalysts per name, including expected product launches, backlog targets, and guidance revisions. This helps you gauge timing and risk tolerance more clearly.

Putting It All Together: A Pragmatic Investment Approach

If the optical networking thesis described by Goldman Sachs and echoed by many market observers holds, investors have three practical levers to participate without overconcentrating in a single name.

• Diversified exposure: Allocate across hardware, components, and network operators to capture different parts of the value chain. This reduces single-stock risk while riding the broader trend.
• Quality over quantity: Favor companies with strong balance sheets, healthy free cash flow, and recurring revenue elements (e.g., software or services alongside hardware).
• Longer time horizon: The AI infrastructure cycle can unfold over several years. Smaller quarterly swings should not derail a well-considered plan to participate in the secular growth story.

Pro Tip: Consider a phased entry approach. Start with a position that aligns with your risk tolerance, then add on stronger signals like improving margins or accelerating backlog in subsequent quarters.

Conclusion: A Measured Path Into a Promising AI Infrastructure Megatrend

The case for optical networking as a megatrend in AI infrastructure rests on three pillars: the need to move ever-larger data sets quickly, the capacity upgrades already underway in data centers and networks, and the early-mover advantages enjoyed by established players in the space. While no single stock guarantees success, a thoughtfully chosen mix of CIEN, LITE, and NOK can offer exposure to the growth in AI-driven data transport without concentrating risk in a single company. And as goldman sachs says that, the long-term demand picture remains intact: AI workloads require faster, more efficient networks, and optical networking is the backbone that can scale with them.

Frequently Asked Questions

Q1: What exactly is optical networking?

A1: Optical networking uses light signals to transfer data across fiber optic cables. It enables high-speed, low-latency communication between data centers and within large networks, which is essential for AI training and real-time inference.

Q2: Why is AI infrastructure demand rising now?

A2: AI models are growing bigger and more capable. Training them requires massive bandwidth, fast memory, and efficient data movement. As a result, data centers and networks must upgrade to handle larger datasets and faster workloads.

Q3: Which stocks offer the best entry points to this trend?

A3: Three well-positioned names are CIEN (Ciena), LITE (Lumentum), and NOK (Nokia). Each provides exposure to optical networking through different angles—systems, components, and telecom networks—helping balance risk and potential upside.

Q4: How should I assess risk in this space?

A4: Look at capex cycles, backlog, gross margins, and customer concentration. A diversified approach across hardware, components, and services reduces the impact of any single cycle downturn.