Many people only focus on chips and GPUs when looking at semiconductors, but ignore the ceramic materials underlying the packaging.
In the past two years, the destocking of the entire semiconductor industry is coming to an end, and orders for high-end packaging and testing production lines continue to be full. Ceramic powders and ceramic substrates, which were originally inconspicuous, have suddenly become indispensable consumables for power chips and AI optical modules.
After a number of authoritative organizations updated the data, the prosperity of the entire ceramic industry chain has steadily increased. The following is a breakdown of the real logic behind it in simple terms.
1. The industry cycle has bottomed out, and the packaging track is the first to step out of the downward range
Semiconductors are typically cyclical. The previous weakness in consumer electronics has dragged down the inventory of the entire industry chain, and the capacity utilization rate of ordinary production lines in the packaging and testing process has been low for a long time.
Relevant data in May 2026 show that the overall growth rate of the global semiconductor market in 2026 is expected to reach 26%, and the bottom of the cycle is basically confirmed.
The capacity utilization rate of advanced packaging production lines of domestic leading packaging and testing manufacturers has rebounded to more than 90%, and process order schedules such as CoWoS and Chiplet have been postponed to the second half of the year. The industry has officially shifted from low-end overcapacity to high-end production capacity shortage, which is often referred to as the packaging inflection point in the market.
If advanced packaging wants to be put into mass production, ceramic supporting materials with high heat dissipation and high stability cannot be replaced. This is also a prerequisite for the explosion of demand in the ceramic track.
This round of packaging recovery is different from previous rounds. The driving force is no longer mobile phones and computers, but three new directions: AI computing power, automotive power devices, and high-speed optical modules, all of which happen to rely on ceramic packaging accessories.
2. The urgent need for heat dissipation of AI computing power directly drives up the consumption of ceramic materials
The skyrocketing power consumption of AI servers is the core driver of the demand for ceramics. The power consumption of ordinary servers is around 2000W, and the power consumption of high-end AI servers exceeds 10000W. The heat resistance and thermal conductivity of traditional organic substrates cannot adapt to the operating environment of high-power chips.
Relevant platform disassembly reports show that Nvidia’s new generation of single-cabinet MLCC uses 600,000 units, a 25% increase from the previous generation. The core media raw material of MLCC is high-purity ceramic powder.
In comparison, ordinary servers use about 2,000-3,000 MLCCs, and AI servers use ten times that amount. The expansion of computing power hardware directly drives upstream ceramic powder consumption to continue to rise.
From the perspective of material properties, the thermal conductivity of aluminum nitride ceramics is 180-220W/m·K, which is hundreds of times that of conventional epoxy resin sheets. The thermal expansion coefficient is highly consistent with silicon chips. It is the standard base material for high-end GPU and high-speed optical module packaging.
The market size of the data center optical module ceramic substrate + tube casing segment is expected to be 7.3 billion yuan in 2026, with a year-on-year growth rate of over 34%.
The long-term expansion logic of the computing power industry remains unchanged, and the in-demand properties of ceramic thermal management materials will continue to strengthen.
3. The explosion of third-generation automotive semiconductors has opened up the long-term incremental space for ceramics
800V high-voltage platforms for new energy vehicles are rapidly popularizing, and SiC and GaN third-generation power chips are widely installed in vehicles. Such high-voltage devices have higher operating temperatures and greater discharge intensity, so ceramic substrates must be used as packaging carriers.
Relevant domestic data show that the total domestic demand for power semiconductor ceramic materials is equivalent to 1.56 billion standard wafers, a year-on-year increase of 18.2%; among which, the demand for new energy vehicles will increase from 15% to 32% in 2021, becoming the largest source of demand.
The amount of ceramic substrate used in a single high-voltage platform model is about three times higher than that of traditional fuel vehicle power modules. The agency estimates that the domestic automotive ceramic materials market will exceed 6 billion yuan in 2026.
In addition to vehicle electronic control, the localization of power modules for photovoltaic energy storage and industrial control is accelerating, and the production capacity of alumina and aluminum nitride ceramics is also constantly being digested.
The global ceramic packaging substrate market size is estimated to be US$4.78 billion in 2026, with automotive electronics demand accounting for 27%, making it the second largest downstream market after communications.
The global electrification penetration rate is steadily rising, and automotive ceramics will become a long-term and stable basic demand for the industry.
4. Layered dismantling of the industry chain: upstream powder shortage, mid-end substrate differentiation
The entire semiconductor ceramic industry chain is divided into three layers: upstream high-purity ceramic powder, midstream substrate and packaging structural parts, and downstream semiconductor device applications. The prosperity of each link is significantly different.
Upstream alumina and aluminum nitride high-purity powders have the highest technical barriers. The overall domestic semiconductor ceramic market will be US$1.87 billion in 2024, a year-on-year increase of 23.4%. It is expected to exceed US$2.6 billion in 2026, with an average annual compound growth rate of 18%+.
The overseas supply of high-end aluminum nitride powder is concentrated. The global high-purity aluminum nitride powder gap will be 12.8% in 2025. This is also a key reason for the long-term high gross profit of the upstream link. The gross profit margin of the industry’s powder link generally maintains a range of 35% – 45%.
The midstream is divided into two major routes: mid-to-low-end alumina substrates and high-end AMB aluminum nitride substrates: the low-end alumina technology is mature, domestic production capacity is continuously released, and production capacity is gradually involved; high-end AMB substrates are suitable for SiC power devices and AI chips, with high technical barriers and long customer verification cycles. The localization rate is less than 30%, and supply and demand have been tight for a long time.
Data for the fourth quarter of 2025 show that the import value of domestic semiconductor-specific ceramic components fell by 8.2% year-on-year, the localization rate increased from 36.5% to 41.8%, and domestic mid-range products gradually realized import substitution.
Downstream is bound to packaging and testing plants, wafer fabs, and chip design companies. Domestic 12-inch wafer fabs will add 240,000 pieces of monthly production capacity in 2025. The purchase volume of ceramic suction cups and cavity accessories inside wafer equipment will increase simultaneously, forming a closed loop of demand transmitted upward from the downstream.
5. Policy + domestic substitution dual support, industry growth certainty increased
Domestic top-level industrial planning continues to focus on electronic ceramic materials. Relevant policies include semiconductor ceramics in the catalog of key choke materials, ceramic substrates are classified as encouraged development categories, and AMB copper-clad laminates are included in the demonstration list of advanced basic materials.
The first and second phases of the big fund are superimposed on local industrial funds, with a total investment of more than 15 billion yuan in the semiconductor ceramics-related industry chain, focusing on supporting breakthroughs in powder formulas and precision sintering processes.
The China Semiconductor Industry Association predicts that the overall localization rate of the industry will exceed 50% in 2026, the localization rate of HTCC and LTCC ceramic packages is expected to exceed 70%, and the localization rate of aluminum nitride substrates will exceed 45%.
Comparing the overseas situation, Japanese companies monopolize high-end aluminum nitride substrates, South Korea is deeply involved in LTCC ceramics for radio frequency, and domestic companies are starting from the mid- to low-end and gradually making breakthroughs upward.
The continued policy support + downstream local chip manufacturers giving priority to domestic verification will continue to accelerate the pace of localization of the ceramic industry chain, which is also the core bonus that distinguishes the sector from traditional cyclical industries.
6. Objectively sort out the existing potential risks in the industry
Hidden risks cannot be ignored in any high-prosperity track. The first is the risk of technological iteration. If new composite heat dissipation materials appear in the future, it will divert the existing market demand for ceramic substrates; secondly, the production capacity of mid-to-low-end alumina substrates will continue to expand, and prices may fall and profits will be compressed in the future.
Secondly, there is the risk of overseas technology blockade. High-end ceramic production equipment and core additives still rely on overseas procurement. Geographical fluctuations may affect the production expansion progress of local companies.
In addition, there are periodic fluctuations in demand for downstream AI and new energy vehicles. If terminal hardware shipments are less than expected, it will reversely suppress the growth of upstream ceramic material orders.
Relevant institutions also pointed out in research reports that from 2026 to 2030, only the high-end segment of the industry will have high growth certainty, and the profit fluctuations of the middle and low-end will increase significantly.
On the whole, the turning point of the packaging cycle, the explosion of AI and automotive demand, and the triple logic of domestic substitution have jointly pushed the semiconductor ceramics industry into an upward cycle. Structural opportunities in subdivisions are prominent, but production capacity differentiation and external risks must also be faced squarely.
Interactive question: When you usually pay attention to the semiconductor segmentation track, do you pay attention to upstream unpopular materials such as ceramics?
Do you think the biggest difficulty restricting the breakthrough of domestic ceramics in the future is technology or customer certification?
Welcome to leave a message in the comment area to exchange opinions.
Warm reminder: All the data in this article are quoted from relevant public research reports. It is only an objective analysis of the industry and does not make any prediction of the rise or fall of the target.
