A $0.50 component is threatening the AI server supply chain. SemiAnalysis reports that multi-layer ceramic capacitors (MLCCs) face an unprecedented price hike and extended lead times as AI server demand surges ~5x by CY27.
Key facts
- MLCCs cost under $1 each per unit.
- Each AI server needs tens of thousands of MLCCs.
- AI server MLCC demand growing ~5x by CY27.
- New MLCC production lines take 2 years to build.
- Lead times extending, prices hiking currently.
A quiet but critical bottleneck is emerging in the AI hardware supply chain. According to @SemiAnalysis_, multi-layer ceramic capacitors (MLCCs) — components that cost under $1 each and regulate voltage stability across every chip in a server rack — are facing unprecedented demand pressure.
The Numbers Behind the Squeeze
Each AI server requires tens of thousands of MLCCs. As hyperscalers and enterprises deploy racks at an accelerating pace, total demand for these capacitors is projected to grow roughly 5x by calendar year 2027. Meanwhile, new MLCC production lines take approximately two years to build and qualify, creating a structural supply lag.
Already, lead times are extending and prices are hiking — the kind of signal that historically precedes allocation and spot-market chaos. Unlike GPUs or HBM, which dominate headlines, MLCCs are a commodity component made by a handful of specialized manufacturers (Murata, Samsung Electro-Mechanics, TDK). Their production capacity is not easily scaled.
Why This Matters More Than It Sounds
The AI server buildout depends on thousands of cheap passives as much as on expensive compute. A shortage of MLCCs — or even a prolonged price increase — adds cost and delay to every rack deployed. The industry has seen this pattern before: in 2018-2019, an MLCC shortage disrupted smartphone and automotive production for quarters.
SemiAnalysis frames this as a spreading bottleneck. The AI supply chain conversation has focused on advanced packaging, HBM memory, and silicon interposers. MLCCs are a reminder that even the most mundane components can become gating factors when demand compounds faster than manufacturing can respond.
What's Not Being Said
The source does not name specific vendors affected, nor does it quantify current lead times or price increases in dollars. The 5x demand growth figure is a projection, not a confirmed order book. Still, the structural logic is sound: MLCC fabs have long construction cycles, and AI server MLCC demand is a relatively new and rapidly growing load on a mature supply base.
What to watch
Watch for quarterly earnings calls from Murata, Samsung Electro-Mechanics, and TDK for explicit MLCC lead time and pricing commentary. Also monitor hyperscaler capex calls for any mention of passive component supply constraints affecting server delivery timelines.
