In 2026, NVIDIA remains the undisputed benchmark in the AI chip sector, but its stock performance is no longer the sole focus within the broader infrastructure space. As of early June 2026, Lumentum Holdings (LITE) had posted a year-to-date gain of approximately 154.48%, while Applied Materials (AMAT) surged about 94.86%, both far outpacing the S&P 500’s roughly 10.35% increase over the same period.
These two companies occupy highly specialized and indispensable positions in the AI compute supply chain—Lumentum in high-speed photonic interconnects, and Applied Materials in semiconductor manufacturing equipment. Lumentum provides the technological foundation for the "optics over copper" transition in AI clusters, while Applied Materials enables the mass production of advanced chips. Notably, neither company directly competes in GPU design; instead, they address critical bottlenecks that arise as AI compute capacity scales.
Data Center Interconnect Bottlenecks: The Structural Inflection Point of Photonics
The expansion of AI clusters presents a fundamental physical challenge—data exchange between GPUs grows exponentially, while traditional copper cable electrical interconnects are hitting physical limits in bandwidth density, transmission distance, and power consumption. According to a Nomura industry report, the 1.6T optical module market is entering an unprecedented "super cycle," with global shipments expected to soar from about 2.5 million units in 2025 to roughly 20 million units in 2026. Meanwhile, silicon photonics penetration in advanced 1.6T modules is projected to reach 50% to 70%.
This technological shift signals a paradigm migration in AI data center network architecture—from primarily electrical interconnects to optical interconnects. GlobalFoundries’ management notes that AI is fundamentally changing data center traffic patterns. The previous user-to-server (north-south) traffic structure is being overtaken by machine-to-machine (east-west) server interconnect traffic. In this context, copper cable connections above 200G per lane face significant challenges in transmission distance, bandwidth density, and power consumption, while optical interconnects offer scalability across all three dimensions.
Growth data from the silicon photonics market further validates this trend. Industry research indicates the global silicon photonics market will reach $1.8–2.84 billion in 2025, rising to $2.3–3.55 billion in 2026, with a compound annual growth rate of about 25.3% from 2026 to 2035. Zooming out to the broader optical interconnect market, leading players like Lumentum and Coherent estimate the AI optics market will hit roughly $90 billion by 2030. This scale marks photonics’ evolution from a "component-level" market to a critical layer in AI infrastructure.
Lumentum: Supply-Demand Imbalance in the Photonics Supply Chain
Lumentum’s growth narrative in fiscal 2026 is well-supported by financial data. For the third quarter ending March 2026, the company reported net revenue of about $808.4 million, up roughly 90% year-over-year. Non-GAAP diluted EPS reached about $2.37, beating market expectations of $2.26. Non-GAAP gross margin expanded sharply to about 47.9%, up 540 basis points quarter-over-quarter, and non-GAAP operating margin hit about 32.2%, up 700 basis points quarter-over-quarter.
This structural growth is driven by two core business segments. In components, revenue was about $533.3 million, up 77.3% year-over-year and accounting for roughly 66% of total revenue. Demand for data center laser chips was particularly strong, with an estimated supply gap exceeding 30%. Narrow linewidth laser component shipments surged about 120% year-over-year, and pump laser shipments grew about 80%. In systems, revenue was about $275.1 million, up 121.1% year-over-year. The 1.6T transceiver entered mass production, and cloud transceiver business grew over 40% quarter-over-quarter.
Lumentum’s long-term growth story rests on two technological directions. First, CPO (Co-Packaged Optics) technology is seen as the next-generation core for AI cluster interconnects, with the company projecting CPO could unlock over $5 billion in incremental revenue opportunities. Second, Lumentum recently acquired its fifth indium phosphide wafer fab, targeting mass production by 2028, reflecting its vertical integration strategy for core photonic devices.
Valuation is a key variable in understanding Lumentum’s current market position. As of early June 2026, the company’s trailing twelve-month P/E ratio was about 165.72, forward P/E about 59.52, and price-to-sales ratio about 30.39. Analyst price targets vary widely, from a conservative $600 to Rosenblatt’s estimate of around $1,300. The company’s Q4 fiscal 2026 revenue guidance is $960–1,010 million, with a midpoint of about $985 million, representing roughly 22% quarter-over-quarter growth. This growth rate will be crucial in supporting future valuations.
Applied Materials: The AI Capex Multiplier on the Equipment Side
Unlike Lumentum’s focus on optical interconnects, Applied Materials operates in the semiconductor manufacturing equipment segment—the upstream backbone of AI compute expansion. Advanced logic chips, high-bandwidth memory, and cutting-edge packaging solutions for AI clusters all depend on capital expenditures in wafer fabrication equipment.
For the second quarter ending April 2026, Applied Materials reported revenue of about $7.91 billion, up 11.41% year-over-year, and trailing twelve-month revenue of about $29.024 billion, up 3.33%. GAAP diluted EPS reached about $3.51, a record high, and non-GAAP gross margin was about 50.0%. The company’s next quarter revenue guidance midpoint is about $8.95 billion, with non-GAAP diluted EPS guidance midpoint at about $3.36.
Management’s qualitative outlook is particularly noteworthy. CEO Gary Dickerson stated that Applied Materials expects its semiconductor equipment business to grow over 30% year-over-year in calendar 2026, significantly outpacing the industry. This outlook is supported by capacity expansion in advanced logic chip manufacturing, DRAM expansion, and rising demand for advanced packaging technologies.
Industry-wide, SEMI data provides macro context. Global spending on 300mm wafer fab equipment is projected to rise about 18% to $133 billion in 2026, and another 14% to $151 billion in 2027, with AI demand as the primary driver. The overall global wafer manufacturing equipment market is expected to reach $126–145 billion in 2026, with wafer fab equipment segment growth of about 9.0% in 2026 and 7.3% in 2027.
Analyst consensus for Applied Materials is "moderate buy." Of 34 covering analysts, 27 rate it a buy and 7 a hold. Institutional price targets range from Morgan Stanley’s $502 to several firms’ $575, reflecting strong consensus on growth certainty during the 2026–2027 equipment capex cycle.
Risks to watch include geopolitical factors and cyclical volatility. In Q2 fiscal 2026, equipment purchases in China still accounted for a significant share of revenue. If export controls tighten further or domestic substitution accelerates, the company’s growth trajectory could be affected. Additionally, the semiconductor equipment sector is highly cyclical. While AI demand has offset the decline in traditional consumer electronics in the short term, inventory cycle changes require ongoing monitoring.
Investment Perspectives: From Semiconductor Equipment to Photonics
Examining Lumentum and Applied Materials side by side reveals two distinct AI investment paths, united by the logic that "compute expansion relies on every layer of infrastructure."
Lumentum addresses the efficiency bottleneck in AI cluster "internal communication." As GPU counts per cluster scale from tens of thousands to hundreds of thousands or even millions, inter-rack and intra-rack data exchange bandwidth becomes a critical constraint on overall computational efficiency. Photonic interconnects directly solve this bottleneck, tightly linking Lumentum’s performance elasticity to AI cluster interconnect bandwidth demand.
Applied Materials tackles the "front-end capacity" bottleneck in semiconductor manufacturing. Every advanced process chip, from design to mass production, undergoes multiple rounds of wafer fabrication, deposition, etching, and inspection. Applied Materials’ equipment spans the entire workflow—from logic chips to memory chips to advanced packaging—making its performance elasticity directly tied to global semiconductor manufacturing investment intensity.
From a valuation perspective, Lumentum is currently in a phase of rapid performance growth and elevated valuation, with the market pricing both its current execution and the technical prospects of CPO and indium phosphide capacity expansion over the next two years. Applied Materials, meanwhile, is experiencing steady but more moderate growth, driven primarily by AI-led equipment capex upgrades rather than a single technology’s penetration leap.
For investors seeking diversified exposure to AI infrastructure, including both companies in their investment universe helps build a multi-layered asset portfolio within the AI compute theme—"from chips to clusters, from manufacturing to interconnects."
Gate Stock Trading: One-Stop AI Infrastructure Allocation
In June 2026, Gate officially launched real stock trading services, enabling direct participation in the US stock and ETF markets using USDT. This feature significantly lowers the barriers between traditional financial markets and crypto assets.
Gate currently supports over 10,000 stocks and ETF assets, covering major US exchanges and liquidity networks including the NYSE, Nasdaq, NYSE Arca, NYSE American, and BATS. Investors can use USDT or USDC as funding sources in their Gate accounts, directly allocating US stock assets from their crypto asset accounts.
From a cost perspective, Gate spot stock trading involves no funding rates, swap fees, or overnight holding charges, reducing the barriers and costs for long-term stock allocation. The current product supports intraday market trading, with plans to gradually expand to 24/7 trading.
For Lumentum (LITE) and Applied Materials (AMAT), Gate stock trading offers an efficient pathway to participate in AI infrastructure investment directly within the crypto ecosystem. Investors can diversify crypto and US stock assets within the same account, eliminating the need to transfer funds between traditional brokers and crypto exchanges.
Importantly, Gate’s real stock trading operates via partner brokers, with actual underlying stock holdings. Orders can be executed on major exchanges such as NYSE and Nasdaq, rather than being tied to tokenized derivatives tracking stock prices. This means shareholders’ rights are legally equivalent to holding stock assets directly at traditional brokers, with dividends, splits, and other corporate actions automatically processed.
Conclusion
In 2026, the AI infrastructure investment narrative is evolving from a singular focus on chip performance to a more layered analysis of the industry chain. Lumentum and Applied Materials represent two critical yet often overlooked dimensions in AI compute expansion: data transmission efficiency and manufacturing capacity supply.
Lumentum’s performance growth validates the commercial impact of the fundamental constraint, "the greater the compute, the more expensive the interconnect." Applied Materials exemplifies the capital expenditure principle, "every process upgrade requires equipment first." Both cases point to the conclusion that when a technology path becomes the focal point of large-scale investment in a very short time, changes in its upstream and downstream supply-demand structure often present greater structural opportunities than the focal point itself.
For investors, including these "AI beneficiaries" in overall asset allocation helps supplement risk exposure beyond the traditional chip-centric AI portfolio. Gate’s convenient stock trading pathway further reduces the conversion cost from crypto to global equity assets, making diversified AI infrastructure allocation a more accessible strategy.
