This note separates two semiconductor capacity problems: where process knowledge is developed and how fabrication work is fulfilled.

The proposal was a centralized technical forum around TSMC/ASML-level process expertise, paired with a distributed network of local fabs on older nodes such as 250 nm to 130 nm.

Semiconductor capacity noteOriginally posted 2022-01-23; expanded here around coordination and distributed fabrication.

Related: The Missing Maker Fab

Article focus: process knowledge, older-node fabs, distributed order fulfillment, capacity, and local manufacturing access.

The Split

Advanced process knowledge benefits from concentration: shared tools, shared metrology, shared tacit knowledge, and people who can compare failures across many runs.

Order fulfillment can benefit from distribution. Older-node local fabs could serve regional demand, education, prototyping, and practical devices without pretending to replace the leading edge.

The Capacity Claim

The point is capacity, not a slogan about democratization. A distributed network of 250 nm to 130 nm fabs would still need disciplined process control, standards, test, and shared knowledge. The useful question is which device classes are valuable at those nodes and can be manufactured closer to where they are used.