The crucial point of optical modules has been jammed.

As demand for 800G/1.6T optical modules surges amid the wave of AI computing infrastructure construction, the compound semiconductor indium phosphide (InP), used as the core substrate for manufacturing optical chips, is evolving from a niche material in specialized fields into a strategic resource for the entire digital economy.

Indium phosphide is currently the only semiconductor that can simultaneously meet four critical conditions: direct bandgap (high electro-optical conversion efficiency), precise wavelength matching (1310/1550nm — the golden window for lowest fiber loss), ultra-high electron mobility (supporting signals above 100GHz), and natural lattice matching with epitaxial materials (enabling the integration of lasers, modulators, and detectors on the same substrate).

This makes indium phosphide difficult to replace in optical communications. Once considered a niche compound semiconductor, it is now moving from behind the scenes to center stage. From price doubling to capacity ramping, from NVIDIA prepaying several billion dollars to lock in capacity to domestic companies achieving full-chain localization of 6-inch wafers, the indium phosphide industry is accelerating expansion.

01 Supply Falls Short of Demand, Prices Skyrocket

Indium phosphide is widely used in DFB lasers, EML lasers, and photodetectors, making it an essential raw material for 800G/1.6T and even the next-generation 3.2T optical modules. Data shows that global demand for indium phosphide substrates is expected to reach 2.6–3 million wafers in 2026, while effective compliant capacity is only about 750k wafers, creating a supply-demand gap exceeding 70%.

This imbalance is directly reflected in prices.

As of April 2026, 2-inch optical communication-grade indium phosphide substrates have surged from $800 per wafer in early 2025 to $2,300–$2,500 per wafer, an increase of nearly 2 times; the price of 6-inch high-end substrates has risen from $1,400 per wafer to $5,000 per wafer, an increase of over 250%.

The fundamental reason for the price surge is the long capacity expansion cycle. From crystal growth furnace construction to customer certification, the entire expansion cycle takes 18–24 months, compounded by reliance on imported core equipment, so capacity release cannot keep up with the steep rise in demand.

In addition to demand, the price increase of indium phosphide substrates is also related to upstream raw materials.

The core raw material for indium phosphide is the rare metal indium. The latest data from China Silver Network (as of July 6) shows that the price of indium metal has reached 5,560 yuan per kilogram, doubling from early 2025 and hitting a nearly ten-year high.

Indium rarely forms independent deposits in nature; most indium is extracted as a byproduct of other metal smelting processes, naturally limiting supply elasticity. Shenwan Hongyuan estimates that by 2027, the indium phosphide sector will drive indium demand by 6.77%. Although the proportion seems modest, it is enough to trigger sharp price fluctuations. The cost curve for indium phosphide substrates is firmly locked at high levels, with limited room for price decline.

More critically, the global indium phosphide supply chain has begun to fracture.

In January 2026, China's Ministry of Commerce announced a comprehensive ban on the export of dual-use items (including InP, indium, gallium, and germanium) to Japanese military users and purposes, while civilian exports require strict licensing and end-user review. Market feedback shows that the rejection rate for Japanese and American companies applying for China-produced indium phosphide substrates has exceeded 80%. Meanwhile, the U.S. Department of Commerce had already initiated anti-dumping and countervailing duty investigations on China's active anode materials in January 2025.

Although no separate tariffs have been directly imposed on indium phosphide, the cumulative effect of export control policies is obvious. The EU, under the framework of the Critical Raw Materials Act, has introduced amendments to reduce excessive dependence on a single country (especially China) and has included recycled content requirements in mandatory standards.

This means that in the future, using indium produced in China not only faces higher compliance costs and export control uncertainties but may also be excluded from some high-end supply chains. All these factors are affecting the global supply and expansion pace of indium phosphide.

02 Downstream Giants Begin Locking in Capacity

As indium phosphide supply becomes a bottleneck for the entire AI computing infrastructure, downstream giants are breaking traditional supply chain boundaries and directly injecting capital upstream.

As early as March 2026, NVIDIA announced it would invest $2 billion each in Coherent and another photonics company, coupled with long-term large-volume procurement agreements, to lock in stable capacity for indium phosphide optical chips for the next several years.

Lumentum's CEO disclosed that EML laser production has increased eightfold over the past three years, yet shipments are still 25%–30% below market demand. In June 2026, Jensen Huang personally attended the groundbreaking ceremony for Coherent's world-first 6-inch indium phosphide wafer fab expansion project. NVIDIA's intention is very clear: in the AI arms race, upstream indium phosphide capacity has become a hard constraint for optical interconnects; without locked-in capacity, it cannot guarantee the delivery of its own AI servers. This "giant direct investment" model is reshaping traditional supply chain relationships, transforming indium phosphide from a general material into a strategically bound resource. It also gives downstream companies the determination to scale up.

On the domestic front, Huawei's subsidiary Hubble Technology invested in Yunnan Germanium's subsidiary Xinyao Semiconductor in 2020, holding 23.91% and becoming the second-largest shareholder.

This investment not only provided financial support but also stipulated in an agreement that Xinyao Semiconductor must prioritize supplying gallium arsenide (GaAs) and indium phosphide (InP) substrate materials to Huawei's affiliates. The collaboration focuses on core materials such as indium phosphide substrates. Xinyao Semiconductor's products have passed Huawei HiSilicon's testing and verification and are used in 5G, data centers, and other fields. In 2025, Huawei secured an order of 80k indium phosphide wafers from Xinyao Semiconductor (53% of capacity), with a prepayment ratio of 40% (industry convention <20%). This investment not only provided financial support but also ensured priority supply rights through the agreement, deepening mutual interests.

03 Global Companies Begin Expanding Capacity

Facing a historic supply gap, major global manufacturers have launched aggressive expansion plans.

Overseas, traditional giants are accelerating their layouts. AXT plans to expand 200 4-inch single-crystal furnaces, with a capacity target of 50k wafers per month in 2026, and aims to quadruple total capacity by the end of 2027; Sumitomo Electric plans to invest approximately 18 billion yen, aiming to increase indium phosphide substrate production capacity to 3.1 times the fiscal year 2024 level by fiscal year 2028; Lumentum expects EML capacity to grow over 50% from 2025 by the end of fiscal 2026, having already advanced about 40% of its indium phosphide (InP) expansion plan; Coherent is expanding 6-inch indium phosphide (InP) wafer capacity in Sherman, Texas, expecting to achieve the target of doubling capacity by the end of 2026 one quarter ahead of schedule, and will double it again by the end of 2027.

Domestic companies are also expanding rapidly.

Yunnan Germanium (through subsidiary Xinyao Semiconductor) is the absolute leader, with existing capacity of 150k wafers per year (converted to 4-inch equivalent). In April 2026, it launched an expansion project with a total investment of 189 million yuan, adding a production line with an annual capacity of 300k wafers (converted to 4-inch, including 6,000 6-inch wafers), bringing total capacity to 450k wafers per year.

Youyan New Materials has existing indium phosphide production capacity of 150k wafers per year (covering 2-6 inch full specifications). Its 6-inch products have completed technical breakthroughs and achieved small-volume supply, with continuously improving yield. The company plans to add 250k wafers per year of indium phosphide capacity, expected to reach production in the second half of 2027, with a total capacity target of 400k wafers per year.

Vital Advanced Materials plans to invest 1.7 billion yuan in fixed assets, using existing sites to upgrade and expand production lines, introducing core production equipment such as high-end crystal growth, precision polishing, and defect inspection, with a focus on 4-6 inch high-end gallium arsenide and indium phosphide single-crystal substrate products. Upon completion, the project will have an annual production capacity of 3 million gallium arsenide substrates and 3 million indium phosphide substrates, totaling 6 million high-end semiconductor substrates per year. The construction period is from August 2026 to August 2029.

Guangdong Pingrui Jingxin's semiconductor technology industrial park has a total investment of 1.1 billion yuan. Upon completion, it is expected to achieve an annual production capacity of 300k indium phosphide single-crystal substrate wafers, with total annual sales revenue exceeding 600 million yuan.

In addition, Sanan Optoelectronics' Wuhan base, China's first 6-inch InP epitaxy mass production line, has expanded its core epitaxial process capacity to 6,000 wafers per month. Xianrui Technology has launched an expansion project for an annual output of 40 tons of indium phosphide crystals, which received environmental impact assessment approval on March 18, 2026 (Qing Gao Approval Environment [2026] No. 3), leaving only the final step before production.

Dingtai Xinyuan is actively expanding its indium phosphide substrate capacity, but the timeline for expansion and reaching full capacity remains uncertain. However, expansion is not achieved overnight. Long production line construction cycles, delivery times of 12–24 months for core equipment such as MOCVD, and customer certification cycles typically requiring 1–2 years mean that the industry's supply-demand tightness will last at least until 2028.

The heat has also attracted cross-sector players. On June 21, 2026, Xingye Technology, a company primarily engaged in natural cowhide leather, announced it would acquire the indium phosphide substrate and semiconductor electronic materials business of Qingdao Li'ang Jingdian for 55 million yuan in cash, covering all assets, business teams, patents, trademarks, and proprietary technologies.

Suqian Liansheng announced in June 2026 that it would enter the indium phosphide substrate track, planning to establish a joint venture company with an initial investment of 100 million yuan to build a production line with an annual capacity of 120k 4-6 inch wafers, expanding to 400k wafers per year in the second phase.

04 Domestic Indium Phosphide Technology Breakthroughs

Beyond capacity expansion, systematic breakthroughs in domestic indium phosphide technology are also noteworthy.

Full-chain localization of 6-inch wafers is the most milestone achievement.

In August 2025, the Jiufengshan Laboratory, in collaboration with Yunnan Xinyao, leveraged domestic MOCVD equipment and InP substrate technology to overcome the challenge of large-size epitaxial uniformity control, successfully developing epitaxial growth processes for 6-inch indium phosphide (InP)-based PIN structure detectors and FP structure lasers for the first time. Key performance indicators reached internationally leading levels.

This achievement also marks the first time in China that domestic collaboration from core equipment to key materials has been realized in the field of large-size indium phosphide material preparation, providing important support for the industrialization of optoelectronic devices.

In terms of crystal growth process innovation, domestic companies are upgrading from traditional LEC (Liquid Encapsulated Czochralski) to VGF (Vertical Gradient Freeze) method. In the past, the mainstream domestic preparation method for indium phosphide had high process difficulty, high dislocation density, and was prone to twinning.

Huaxin Jingdian uses the vertical gradient freeze (VGF) method to prepare indium phosphide single crystals, resulting in higher product quality and stability. Vital Advanced Materials has independently developed VGF indium phosphide single-crystal growth technology, combined with low-damage wafer polishing and ultra-clean surface cleaning key technologies, successfully producing 6-inch indium phosphide substrates with low dislocation density, stable electrical properties, high flatness, and clean surfaces.

Heterogeneous integration is also advancing simultaneously. Hybrid/heterogeneous integration of InP and silicon photonics (SiPh) is the mainstream technical direction in the current optical communications field.

InP provides light sources (lasers, amplifiers), while silicon provides passive waveguides and electrical interconnects. The two are integrated through wafer bonding, micro-transfer printing, or 3D hybrid integration to achieve optoelectronic integration. Commercial optical transceiver modules from Intel and Cisco use heterogeneous integration technology. In China, Jiufengshan Laboratory and Sun Yat-sen University have also successfully achieved heterogeneous integration of indium phosphide lasers on silicon wafers, demonstrating the feasibility of large-scale production.

05 Conclusion

Looking back from mid-2026, the surge in indium phosphide is not simply a cyclical shortage but a violent collision between the AI computing revolution and the semiconductor materials supply chain.

Just in early July, Huawei's He Tingbo released an updated version of the "Time Scaling Theory of Multilayer Electronic Systems" V2. Tao's Law 2.0 defines τ (time constant) as a layered composite variable spanning four levels: devices, circuits, chips, and systems. Its value is jointly determined by underlying hardware parameters, the tier architecture, and communication overhead.

If LogicFolding takes shortcuts for signals at the circuit layer and uses 3D stacking at the chip layer to compress routing delays, then system-level τ optimization points to a harsher reality: over 80% of energy consumption in large AI clusters is spent on data movement; over 70% of system cost is allocated to data storage. The direct inference is that reducing the transmission time of data between chips, racks, and packages is no less important than shortening the computation time itself.

This is precisely the strategic significance of indium phosphide. Huawei's Hi-ONE high-density optical interconnect engine and unified memory semantic bus (Lingqu Bus) deployed at the system level aim to push cross-rack optical interconnect bandwidth to 8 Tb/s per lane and compress SerDes transmission distance from 100 cm to 5 cm. The realization of these system-level τ compression all depends on indium phosphide optical chips.

Source: Semi-Transverse

Risk Warning and Disclaimer

        Market risks exist; investment needs caution. This article does not constitute personal investment advice and does not consider the special investment objectives, financial situations, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Investments made based on this are at your own risk.