Lead Times for Key Chip Equipment Components Stretch Up to 40 Months
nashnova research
Lead times for critical semiconductor-equipment components have stretched to as long as 40 months, far outpacing equipment makers' own capacity ramp. This means → the bottleneck for new fab capacity has shifted from "can you buy the machine" to "can the machine get its parts."
How much have parts lead times actually stretched?
A Korean deposition-equipment maker says components that once arrived in roughly 4 months now take 10 months — more than double.
The extreme case: a Korean laser-processing equipment firm reports certain critical Japanese-made parts now carry a 40-month wait, over three years.
Packaging-equipment suppliers see a milder squeeze — local-sourced parts have gone from a typical 4 months to over 6 months, roughly a 50% increase.
These are individual company reports, not industry averages, but they point to a systemic constraint forming upstream of major equipment vendors.
Why can't parts suppliers keep up?
The core issue: component makers did not expand capacity ahead of the equipment-demand surge, leaving insufficient reserves.
This reflects a hedging mindset upstream — some Japanese suppliers doubt whether this investment cycle will last, and are reluctant to commit to new capacity.
In plain terms = equipment makers are scrambling to fill orders, but the firms making the parts inside those machines fear the boom will fade before the investment pays off.
How fast is TSMC's equipment demand growing?
TSMC COO Y.J. Mii disclosed at SEMICON Taiwan 2026 that the company's equipment-need estimate for 2026, indexed to 1.0 at end-2025, rose to 1.5× after Q1 and reached 1.9× by July — nearly doubling in six months.
TSMC is simultaneously building close to 20 fabs worldwide: 13 in Taiwan, five to six overseas. Previously, the company managed only four to five concurrent projects.
This means → equipment demand is not growing linearly; it jumped in a matter of months, leaving almost no buffer for the supply chain.
How are ASML and Applied Materials responding?
ASML's EUV lithography systems — machines that etch chip circuits with extreme-ultraviolet light — cost roughly $200 million each. Capacity through 2027 is nearly sold out; the company is studying ways to expand EUV output further.
Applied Materials plans to double its semiconductor-systems quarterly output by 2028, with options to scale further before 2030.
But equipment makers can only ramp as fast as their parts suppliers do — final assembly is quick, while specialty components and subsystems have far longer expansion cycles.
What do the industry forecasts say?
SEMI projects global semiconductor equipment sales will rise 23.2% year-on-year in 2026, hitting a record $165.9 billion, then climbing further to $229.5 billion by 2028.
Wafer-fab equipment is forecast to grow 23.1% to $143.9 billion. DRAM equipment spending, driven by HBM — high-bandwidth memory, the fast memory stacked on AI chips — is expected to jump 39% to $38.8 billion.
This means → the money on the demand side is already committed, but whether it converts into actual capacity depends on how much can flow through the narrowest pipe: component supply.
What is the real risk for chipmakers?
Capital expenditure can be booked quickly, but capacity delivery depends on equipment arriving on time — and parts delays inside that equipment add another layer of uncertainty.
In plain terms = placing an order does not mean receiving a machine, and receiving a machine does not mean every critical component inside it is ready. The bottleneck nests within the bottleneck.
This reflects a shift in competitive logic: securing production slots in the supply chain, not merely placing purchase orders, may become the variable that determines when new fabs actually come online.
市场有风险，内容仅供研究参考，不构成投资建议。