Test report DSG-1819 · Rev B · tested October 12, 2026

Supply Chain & PolicyDevice under test

ASML Sees AI Chip Boom Running Past 2028; Broadcom Flags Financing Risk

ASML's CEO says AI chip demand could run past 2028; Broadcom counters that financing, not demand, may cap the industry's expansion. Here's what each position implies.

Read
2 min
Words
480
Node
7nm
Operator
Grace Kim

Spec summary

  1. ASML's CEO says the AI chip boom could stretch beyond 2028
  2. Broadcom warns that financing may become the industry's bottleneck
  3. ASML holds a monopoly on EUV lithography, the tooling required for advanced chips
  4. Broadcom designs custom AI accelerators for large cloud operators
ASML CEO Says AI Chip Boom Could Stretch Beyond 2028, but Broadcom Warns of Financing Bottleneck - Yahoo Finance
Fig. AASML CEO Says AI Chip Boom Could Stretch Beyond 2028, but Broadcom Warns of Financing Bottleneck - Yahoo Finance — AI-generated

ASML's chief executive says the AI chip boom could stretch beyond 2028, but Broadcom has countered with a warning that financing, not demand, may become the industry's binding constraint.

The two positions, reported by Yahoo Finance, frame the central question for the semiconductor supply chain over the next four years: how long the current build-out lasts, and who can actually pay for it.

ASML, the Dutch company that holds a monopoly on extreme ultraviolet lithography (EUV — the machine technology required to print the most advanced chip circuitry), sits at the front of every capacity decision in the industry. When its CEO puts a horizon like "beyond 2028" on the AI demand cycle, that is a supply-chain signal, not a marketing line. Chipmakers commit to EUV tool orders years ahead of delivery, so ASML's order book functions as a leading indicator (a metric that moves before the broader market does) for fab construction and chip output.

What does this change for buyers and builders?

If ASML's read holds, the answer is: planning horizons extend. Foundries, memory makers and hyperscalers (the large cloud operators that buy AI accelerators in volume) would face pressure to lock in capacity well into the decade's back half. That supports continued heavy capital spending across the equipment ecosystem — lithography, deposition, metrology, advanced packaging.

Broadcom's warning cuts the other way. The company, which designs custom AI chips for large cloud customers, says the bottleneck could be financing. The concern is straightforward arithmetic: AI data-center build-outs now require tens of billions of dollars per project cycle, and the pool of players able to fund that at scale is narrow.

The tension is real. Demand signals and funding capacity are diverging.

Two views, one supply chain

  • ASML: AI-driven demand could persist beyond 2028, implying a longer equipment cycle than skeptics assume.
  • Broadcom: Access to financing may cap how fast that demand can actually be served, regardless of technical readiness.

Both companies occupy leverage points in the AI silicon stack. ASML sells the tools nobody can bypass. Broadcom designs the custom accelerators that hyperscalers deploy instead of, or alongside, merchant silicon.

Why the financing argument matters

Chip demand can outrun the balance sheets meant to fund it. If capital markets tighten, or if AI project returns disappoint investors, expansion plans slow regardless of end-market appetite. Broadcom's caution effectively asks whether the industry's growth curve is demand-limited or capital-limited.

For equipment suppliers, the distinction is everything. A demand-limited boom rewards capacity expansion. A capital-limited one rewards balance-sheet strength and customer concentration among the few who can still write the checks.

Watch ASML's quarterly bookings and any broadening of Broadcom's customer base beyond its largest cloud clients. Those two data points will show which CEO's framing the market is pricing in.

via Google News: AI chip (Source)

Filed under

  • asml
  • broadcom
  • euv-lithography
  • ai-accelerators
  • semiconductor-capex
Share this article:

More from Grace Kim

Grace Kim

Show full bio

Market editor covering regulation, policy and their impact on hardware layer of artificial intelligence at Die Signal.

282 articles

Same lot · LOT-C1C8

« Previous article