Physix FrontierPhysix Frontier · News Briefing Card (enbrief · Oct 23, 2026)
Intel CTO: Leading-Edge Wafer Output Can At Most Double
KEY FACTS
- Intel CTO Ranade said on a podcast that the upper limit for growth in global leading-edge wafer output is roughly two times.
- He said this constraint comes from the semiconductor supply chain's overall ability to coordinate capacity expansion, not from any single link.
- Ranade said AI compute has grown roughly threefold per year over the past decade.
- ASML produces fewer than 100 EUV lithography machines per year, and significant capacity expansion will be difficult even by the end of this decade.
- Equipment is delivered about 18 months after orders are placed, and new capacity will only come online around 2028.
KEY DATA
2xLeading-edge wafer output growth ceiling
about 3xAI compute annual growth rate
fewer than 100 unitsASML EUV lithography machine annual output
about 18 monthsEquipment delivery cycle
PHYSIX OBSERVATION
AI compute demand triples every year, while wafer capacity can only double, so the gap will persist for a long time. The bottleneck is not money, but the synchronized expansion of the entire chain - lithography machines, components, materials - and any link falling behind drags down the whole. This means the compute shortage is not a short-term phenomenon, and chip price increases and the battle for capacity may continue past 2028.
Source: enbrief original report ↗