Ajinomoto's core strategy for its critical ABF material has become clear: prioritize stable supply over short-term profits.
At the heart of today's most powerful AI chips, from GPUs to server CPUs, lies a special insulating film called Ajinomoto Build-up Film (ABF). As AI models become more complex, the chips that run them get bigger and more intricate. This requires advanced packaging techniques like TSMC's CoWoS, which stack multiple chips and memory modules (like HBM) together. This process consumes a large amount of ABF, turning this once-obscure material into a critical bottleneck for the entire AI industry. If you can't get enough ABF, you can't build enough AI accelerators, even if you have the silicon wafers ready.
Recognizing this, Ajinomoto has made a strategic choice. The company’s CEO recently confirmed that they will focus on expanding production to meet demand through 2030, rather than exploiting the shortage with opportunistic price hikes. First, the company has already secured land for a new factory in Gifu, Japan, which is slated to begin operations in 2032, specifically to meet future AI-related demand. This move directly addresses the supply crunch worries.
Second, this strategy stands in contrast to pressures from both the market and investors. While suppliers of related materials like copper-clad laminates (CCLs) are raising prices due to surging raw material costs, Ajinomoto is holding firm. This is despite activist investors pushing for price increases of over 30%. By choosing stability, Ajinomoto is building 'relationship capital' with key clients like TSMC, ensuring it remains an indispensable partner for the long haul.
Finally, while new technologies like glass-core substrates are on the horizon, the industry consensus is that they will coexist with, not replace, ABF-based organic substrates for the foreseeable future. Ajinomoto’s commitment to expanding capacity reinforces the view that ABF will remain the workhorse material for the AI era, making their strategy a pivotal factor in the pace of AI development.
- ABF (Ajinomoto Build-up Film): A resin-based insulating film used in the substrate of high-performance semiconductor packages. It is essential for creating the fine circuit patterns needed for chips like CPUs and GPUs.
- CoWoS (Chip-on-Wafer-on-Substrate): An advanced packaging technology developed by TSMC. It allows multiple chips and high-bandwidth memory (HBM) to be integrated on a single silicon interposer, improving performance and efficiency.
- HBM (High Bandwidth Memory): A type of high-performance RAM that stacks memory chips vertically to provide much faster data transfer speeds compared to traditional memory, crucial for AI applications.
