Ajinomoto Reportedly Eyes 30% ABF Film Price Hike as AI Server Demand Keeps Substrate Market Tight Through 2027
The AI infrastructure boom is now pushing pressure deeper into the semiconductor supply chain, and the latest market reports suggest ABF substrates are becoming one of the next major bottlenecks. According to a new report citing Taiwan supply chain sources, Ajinomoto, the company best known globally for MSG but also the dominant force behind Ajinomoto Build up Film, is reportedly preparing to raise ABF film pricing by at least 30% as supply demand imbalances stretch further into 2027. The development points to a fresh round of cost pressure for AI GPUs, AI ASICs, and the broader advanced packaging ecosystem.
ABF, or Ajinomoto Build up Film, is a critical insulating material used in advanced semiconductor packaging and has become a de facto standard for high performance chips. Ajinomoto itself describes ABF as a core material for high performance semiconductor substrates and says demand is expected to keep growing as AI, cloud, networking, and other high performance computing applications become more complex and multilayered. In practical terms, that means ABF is not just another material in the stack. It is part of the backbone connecting next generation compute silicon to the substrate layer used in many of today’s most advanced processors and accelerators.
The new report says the supply demand gap for ABF substrates is now expected to continue through the end of 2027, with Taiwanese substrate makers already running at full utilization. That is a notable signal because it suggests this is no longer a short cycle squeeze tied to a single launch window. Instead, the market appears to be dealing with a broader structural shortage driven by sustained AI server deployment, rising demand for advanced packaging, and limited ability to expand supply fast enough to fully catch up. Taiwan market reporting from this week similarly says the ABF imbalance is expected to persist through 2027, supporting stronger pricing conditions for major substrate suppliers.
According to the report, contract pricing for ABF substrates is expected to rise by around 5% to 10% in the second half of 2026, while spot pricing has already moved much more aggressively, reportedly climbing more than 30%. The bigger flashpoint, however, is the claim that Ajinomoto may raise ABF film pricing by at least 30%. If that happens, the cost increase would likely ripple through substrate makers and eventually into end customers, particularly those buying packaging for AI accelerators and custom ASIC deployments.
The reason this matters so much is simple. In the AI era, packaging is no longer a background cost. It is one of the strategic choke points in the entire compute supply chain. As chip complexity rises and server platforms demand more advanced substrates, even a relatively narrow materials shortage can create wider pricing consequences across the board. Market commentary cited in recent coverage suggests that a 30% increase in ABF film prices would translate into roughly a 3% to 6% increase in final substrate pricing for AI GPUs and ASICs. That may sound manageable on paper, but at hyperscale server volumes it becomes a meaningful cost lever.
This is also turning into a margin story for the substrate industry itself. With utilization tight and pricing improving, top Taiwanese ABF substrate makers such as Unimicron, Kinsus, and Nan Ya PCB are expected to benefit from stronger revenue and profitability. Some reports now point to gross margin expectations of 22% to 30% in 2026 and 30% to 35% in 2027, assuming demand conditions remain firm and pricing discipline holds. That outlook reinforces how the AI supercycle is not only lifting GPU and server vendors, but also reshaping the economics of the materials and packaging companies deeper in the chain.
Looking ahead, the pressure may not ease quickly. Ajinomoto has already signaled that ABF demand should continue growing in line with data center, AI, and networking trends, and the company is expanding production for future semiconductor market growth beyond 2030. Even so, near term supply appears likely to remain constrained, especially as next generation AI servers and emerging technologies such as co packaged optics add more demand across advanced packaging layers. For the broader semiconductor market, this is another reminder that the next AI bottleneck may not come from the GPU die itself, but from the substrate and materials ecosystem that makes large scale deployment possible.
Do you think ABF substrate constraints could become one of the biggest hidden cost drivers in the AI server race over the next 2 years?
