What happened
Osaka-based Sakai Chemical, known for fine powders used in geisha face makeup, now supplies raw materials for capacitors in AI data centers. Its production capacity is quickly filling up, and the company is calculating how much more to add, per President Toshiyuki Yagura.
Why it matters
A century-old cosmetics-powder maker is turning into a little-known AI supply chain linchpin, with the last year transforming Sakai Chemical, suggesting niche materials suppliers may sit on a critical path for AI infrastructure buildout.
What to watch
The company's expansion plans hinge on how much extra production capability it decides to add, and any capacity steps it announces will be the figure to watch.
WHO IT HITSThis lands on procurement and supply chain managers sourcing materials for servers and data centers, who may need to track capacitor raw-material availability from a supplier that until recently treated it as a side business.
Summaries like this, in your inbox every morning.
Sakai Chemical spent decades perfecting the production of ultra-fine and uniform particles. One outlet for that skill was a safe, lead-free base material for traditional Japanese face powders; another was chemical compounds able to store large electrical charges inside tiny components. The capacitor materials were, until recently, a side business.
What changed is demand from artificial intelligence data centers, which need capacitors in servers. As that demand has scaled, capacity at the Osaka company has quickly filled, and management under President Toshiyuki Yagura is working out how much more production capability it needs to add.
The stakes for Sakai Chemical appear to hinge on how far it commits to expanding capacity and whether that expansion is enough to hold its place in the AI supply chain. For buyers of server components, a maker of cosmetics powders moving onto the critical path is a reminder that AI buildouts depend on less visible suppliers as well.
For example, today's edition would include:
AI-summarized, only the topics you pick: one digest a day via Email, LINE, or Slack.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
