Ajinomoto: The MSG Company Inside Every CPU

1From umami to insulation

Ajinomoto invented monosodium glutamate seasoning in 1909. Ninety years of amino-acid chemistry later, its researchers solved a problem the electronics industry had been fighting through the 1990s: how to insulate the ever-finer wiring layers of a flip-chip package substrate. The answer was an epoxy formulation delivered as a dry film - Ajinomoto Build-up Film - that could be laminated, laser-drilled and plated with far better precision than the liquid resins it replaced.

Intel adopted it around the turn of the century. Within a few years ABF was the de facto standard for high-performance package substrates, and it has never been displaced.

2Why nobody displaced it

ABF does not sit alone - it sits inside a stack that the whole substrate industry co-optimised around it. Ibiden, Shinko, Unimicron, Nan Ya and AT&S tuned their laser drills, desmear chemistry, copper plating and reliability qualifications to ABF's exact behaviour over two decades. A challenger film must not merely match it; it must repay the requalification of every one of those steps at every substrate maker for every chip family.

Competing films exist - Sekisui and Taiyo make credible chemistry - and they remain niche. That is what a process-embedded monopoly looks like: the moat is not the patent, it is everyone else's sunk qualification.

32021 was the proof

Through 2021-22, ABF-class substrates were the binding constraint on CPU and GPU supply. Lead times on high-end packages stretched toward a year; Nvidia, AMD and Intel all shipped less than demand while wafer fabs had capacity to spare. The industry's response is telling: nobody switched films. Everybody - Ajinomoto included - built capacity. When a shortage produces investment in the incumbent rather than substitution, the lock-in is real.

Ajinomoto's own disclosures through the period showed its electronic materials line growing at rates its seasoning business never will, off a base small enough that most shareholders still own it for food.

4The AI kicker

Accelerator packages are physically enormous, and getting bigger: silicon interposers and chiplet assemblies (TSMC's CoWoS and successors) still sit on organic ABF substrates, whose area per device keeps rising. Film demand scales with package area, not die count - a quiet multiplier on every AI capex cycle.

The visible print of Ajinomoto's invisible layer is substrate-maker capex: when Ibiden announces a new plant for Nvidia-class substrates, ABF volume is the assumption underneath it.

5The investable surface, honestly

Ajinomoto trades in Tokyo (2802) with an OTC ADR (AJINY). The majority of its profit remains food and amino sciences; functional materials is the strategic kicker. The levered expressions are the substrate makers - Ibiden in Tokyo, AT&S in Vienna, with Shinko in the process of going private - and the downstream dependents are simply the processor complex: NVDA, AMD, INTC, AVGO, and TSMC as the packaging integrator.

The honest framing: you cannot buy ABF cleanly. You can only know that it is there, under every accelerator earnings call, and that its supply decisions in Kawasaki matter more to GPU availability than most things that get daily coverage.

Sources and method

Compiled from public filings, company disclosures and trade press, cross-checked against our desks where they overlap. Supply shares in private industrial chains are estimates, are labelled as estimates, and the ones we could not verify are listed under what we do not know rather than asserted. The lines below are the factual spine.

• ABF has been the standard build-up dielectric for flip-chip substrates since ~1999 (industry consensus; Intel adoption)
• 2021-22: ABF-class substrate supply gated CPU/GPU shipments while wafer capacity was available (company statements across the complex)
• Package substrate area per AI device is rising with chiplet/interposer designs (TSMC packaging disclosures)

Underlying sources

• Ajinomoto disclosures - functional materials segment
• Substrate-maker capex announcements (Ibiden, AT&S)
• Industry reporting on the 2021-22 ABF/substrate shortage
• Parallax bottlenecks board - advanced packaging layer

Written by Parallax Research's research writer on 2026-07-29 and reviewed against the dossier it was given. Figures are as of that date and are not updated in place. Our scoring model's out-of-sample AUC is published in the whitepaper; nothing here is a forecast.