Ajinomoto Cuts ABF Film Supply to Mainland China by 30% as AI Demand Squeezes Chip Substrate Materials

Ajinomoto, the Japanese group that controls the global market for build-up film used in advanced chip packaging, has told customers in mainland China that it will reduce their allocation by 30 percent. The information was reported on August 13, 2026 by Chinese industry media citing supply chain sources. As of August 15, 2026, no third party has confirmed the notice and Ajinomoto has not issued a public statement on the reported allocation change.

ABF stands for Ajinomoto Build-up Film. It is the insulating layer used in the IC substrates that carry high-end processors, graphics chips, programmable logic devices and custom AI accelerators. The material represents roughly 30 percent of the total cost of a finished substrate, which makes it both a technical and a commercial bottleneck for anyone building advanced packages.

The product came out of research on by-products of the company's seasoning business in the 1990s. Ajinomoto worked with Intel from 1996 and reached mass production in 1999. Three decades of patents and long qualification cycles have left the company with more than 95 percent of the global market, a position that has no equivalent elsewhere in the semiconductor materials chain.

Supply has been tight all year. Monthly output passed 2 million square meters in the second quarter of 2026 and the lines have been running at full load since then. The next expansion phase is not scheduled to start before 2028, and a third plant in Gifu Prefecture is expected to reach production only around 2032. The company has committed about 25 billion yen through 2030 to lift capacity by roughly half, an increase that many analysts consider modest against the growth rate of AI accelerator shipments.

Faced with that ceiling, Ajinomoto appears to be protecting its Japanese customers and its largest overseas accounts, and the mainland Chinese quota is absorbing the reduction. Taiwan accounts for around 42 percent of ABF film consumption, Japan around 27 percent and mainland China around 14 percent, so a cut concentrated on one regional market has a limited effect on total volumes while creating an acute problem for the customers concerned.

The allocation news follows a price move that is already in force. In March 2026, Palliser Capital, one of Ajinomoto's larger shareholders, publicly pushed the company to raise ABF prices by more than 30 percent and to separate its electronic materials business. In May, Taiwanese substrate makers confirmed receiving formal notices of an increase of up to 30 percent, applied customer by customer and effective from the third quarter of 2026. Resonac and Mitsubishi Gas Chemical had already raised copper clad laminate prices by a similar margin earlier in the year, so the whole upstream substrate chain is repricing at once.

Delivery cycles have stretched accordingly. Market sources put current lead times for ABF film at more than six months. Lead times for finished ABF substrates have moved from roughly twelve to sixteen weeks toward twenty to thirty weeks.

Forecasts for the gap between supply and demand keep widening. Goldman Sachs estimated in February 2026 that the shortfall in ABF substrates would reach 10 percent in the second half of 2026, 21 percent in 2027 and 42 percent in 2028. Morgan Stanley has raised its 2030 estimate for the high-end ABF substrate gap from 15 percent to 22 percent. Demand per chip is also rising, because an AI accelerator package uses eight to sixteen build-up layers against four to six for a standard PC processor, on a much larger surface, which multiplies film consumption by a factor of five to ten.

For mainland China the exposure is direct. Local self-sufficiency in ABF film remains below 5 percent. Substrate makers such as Shennan Circuits, Xingsen Technology and Shenghong Electronic depend on imported material for their high-end lines. A 30 percent reduction translates into idle qualified capacity, delayed customer programs and higher unit costs at a moment when domestic AI chip designers are trying to scale packaging volume. Upstream fiberglass cloth is a second constraint, since high grade T-glass supply is also short, so relieving one input does not solve the other.

Domestic suppliers are moving, but the timeline is measured in years rather than quarters. Huazheng New Materials is scaling a composite build-up film branded CBF and has reported yields above 85 percent, against more than 99 percent for the incumbent material. Lianhua Holding is developing an NBF product, Hongchang Electronics is working on a GBF film and Shengyi Technology is pursuing an ABF-like route. Several of these programs have passed the sample stage and entered small volume supply, and new lines are due to start up around the end of 2026. Full qualification at substrate and assembly plants normally takes two to three years, so broad replacement of imported film is unlikely before 2028.

In the short term, Chinese substrate makers are expected to reshuffle order books toward their most profitable programs, draw down inventory and look for material through partners in Taiwan and South Korea. In the medium term, the episode strengthens the case for state backed investment in local build-up materials and for faster validation cycles at packaging and test houses. The wider lesson for buyers outside China is that a single supplier holding more than 95 percent share can reallocate volume between regions at short notice, and that no commercial contract offers real protection when capacity is the binding constraint.