Author: Rita
The tight situation for high-end IC package substrates is creating price rigidity. Japan's Ministry of Economy, Trade and Industry's June data provides the most compelling evidence.
Nomura Securities released an industry update report on August 17th, showing that June shipments of rigid module substrates reached ¥291 billion, a year-on-year surge of 98%. This marks the second consecutive month that the category has set a new historical record. Calculated by area, shipment volume increased by 20%; the average price per square meter rose to ¥1.429 million, a year-on-year increase of 65%. Both indicators hit new highs simultaneously. Nomura judges that the tight supply-demand balance for high-end package substrates will be difficult to alleviate over the next 2 to 3 years.
Low Yield Rates for Large-Sized Substrates Drag Down Effective Supply
Shipment value doubled while area increased only 20%. The contrast between soaring price and limited quantity growth points to the same conclusion: high-end package substrates are experiencing a structural supply bottleneck.
Nomura pointed out in the report that the current tight supply-demand situation is concentrated in high-end substrates used for server CPUs, network switches, AI datacenter GPUs, and custom ASICs. Unlike the previous tightness in 2022, which was driven by demand distortion—where the world's largest CPU manufacturer hoarded substrates amid market share loss, artificially suppressing competitors' production—the current core issue lies on the supply side.
As high-end package substrates migrate to larger sizes, production yield rates have declined significantly. Although shipment area has grown, the expansion of effective capacity is slower than demand growth. Relying solely on existing capacity, the tight supply-demand situation will be difficult to alleviate in the next 2 to 3 years.
The scarcity of advanced packaging production lines further exacerbates the problem. Globally, only a handful of factories can produce high-end package substrates. Priority projects have already filled leading production lines, leaving lower-priority projects to compete for capacity on lines producing previous-generation substrates. Nomura warns that this cascading squeeze may trigger new supply issues at the downstream system integration stage.
New Technology in 2027 Will Compete for Advanced Capacity
The market looks to 2027 for relief from new capacity, but Nomura sees another layer of pressure.
The report highlights a new technology set for mass production—Embedded Multi-die Interconnect Bridge with Through-silicon vias (EMIB-T)—expected to enter mass production in the second half of 2027. EMIB-T has a more complex manufacturing process than ordinary high-end package substrates, with extremely high technical difficulty, and can only be produced on advanced production lines. This means its mass production will further compete for the already tight advanced capacity.
Ibiden is Nomura's preferred pick in the industry, with a Buy rating. The company possesses the industry's largest number of advanced production lines and will be a major supplier for EMIB-T. However, Nomura notes that even though Ibiden has prepared some substrate inventory in advance for Rubin processors, it had to prioritize expanding production of server CPU substrates (ARM core) per client requests from July to September. Increases in Rubin substrate production will have to wait until October to December or later. The fact that even a leading supplier must prioritize among client orders underscores the rigidity on the supply side.
Price Rigidity to Persist, Setting Market Expectations
Following the ¥291 billion in June, there is still room for upward movement. Nomura expects subsequent data to remain high until an inflection point in capacity expansion appears.
The market hopes capacity expansion will solve the problem. Nomura provides a more conservative timeline: after EMIB-T enters mass production in the second half of 2027 and consumes significant advanced capacity, a supply-demand balance may not be achieved until 2028 or even later.
For investors, this means the price rigidity of high-end package substrates will not ease over the next 2 to 3 years. Related suppliers' pricing power and profit margins are expected to remain at historically high levels. The core of industry competition is shifting: from "who can produce it" to "who can produce more," making capacity expansion capability a new competitive dimension.
Nomura specifically points out that due to the high concentration of clients in the package substrate industry, it is difficult for outsiders to accurately assess the true extent of the supply-demand tightness. This tension may only become fully apparent to the outside world when bottlenecks emerge at the integration stage. By then, the market may have missed the earliest pricing window.
High-end package substrates are undergoing a structural supply bottleneck. The AI computing power race is pushing every link of the supply chain to its limits. Package substrates are just the first bottleneck to surface. When EMIB-T begins occupying advanced production lines in the second half of 2027, how long can the tight balance for package substrates hold?
Disclaimer
This article is Tide Research's compilation and interpretation of a third-party securities research report (Nomura Securities, August 17, 2026), combined with publicly available market information. The ratings, target prices, earnings forecasts, and related judgments cited herein are the views of the analyst from that securities firm, representing only the stance of their institution. They do not represent the views of Tide Research nor constitute any investment advice.
The market carries risks, and decisions should be made independently. This article should not be used as a basis for buying or selling any securities.
