Where Chip Capacity Comes From: Fabs, Materials and the 5-Hour Line

"Capacity" sounds like a simple number: how many wafers a plant can push through in a month. It is not. Capacity is a chain of processes, materials and machines, and the weakest link sets the ceiling.

Here is what the record shows about where that ceiling comes from, drawn from reporting by IEEE Spectrum, Tom's Hardware, TechCrunch, CNBC, Texas Instruments and Construction Physics.

The original automated line did it faster

Start with IEEE Spectrum's history of IBM's Project SWIFT. It shows how far the industry has moved away from single-day fabrication. In 1970, Bill Harding, then manager of IBM's Manufacturing Research group in East Fishkill, New York, proposed a fully automated wafer-fabrication line that would turn out integrated circuits in less than one day. At the time, IEEE Spectrum notes, ICs such as random-access memory chips typically took a monthlong stop-and-go march through dozens of manual work stations. The numbers in that account are the ones worth keeping. SWIFT averaged 5 hours to complete each layer of its fabrication process. The fastest modern fabs take 19 hours per processing layer, according to the same article, and the industry average is 36 hours. Modern ICs are built with many more layers, on larger wafers, and the processing is more complex. IEEE Spectrum argues those factors do not altogether close the gap.

Harding's automated manufacturing line was really, truly, swift.

Today's billion-dollar fabs measure the fabrication time of an advanced IC in weeks, not days, IEEE Spectrum reports. The article also notes that SWIFT's innovations are now commonplace in highly automated chip plants, but its turnaround time has never been equaled. That is the useful way to read capacity claims: a fab's output depends on how many layers it can process and how fast, not just how big the building is.

National programs are adding capacity, not layers

Governments now compete on capacity directly. On 15 July 2026, TechCrunch reported that India unveiled a ₹625 billion (about $6.5 billion) Mobile Phone Manufacturing Scheme, a five-year program that rewards smartphone manufacturers on eligible sales with incentives ranging from 2.25% to 5%, plus an additional 1.5% for sourcing key components and sub-assemblies in India.

New Delhi also committed a further ₹1.28 trillion (around $13.3 billion) to bolster domestic semiconductor manufacturing, expanding a $10 billion chip incentive program launched in 2021. The same TechCrunch piece shows the scale of the gap. China accounted for 63% of global smartphone production in 2025, compared with India's 18%, according to Counterpoint Research. India's smartphone program runs through March 2031, and the government expects mobile-phone production during that period to total about ₹39 trillion (around $405 billion) and the scheme to create about 60,000 direct jobs. Apple began assembling iPhones in India in 2017 and about 25% of its iPhones are now made there, TechCrunch reports.

Malaysia is taking a different route into the same chain. CNBC reported on 18 April 2024 that Malaysia holds 13% of the global market for chip packaging, assembly and testing services, according to a Malaysian Investment Development Authority report dated 18 February. Second finance minister Amir Hamzah Azizan told CNBC's Karen Tso at the IMF spring meetings in Washington that "the semiconductor space is now in the upcycle — and Malaysia will be a beneficiary of that."

Asked about competition from China and the US, Amir Hamzah said the country does not see China, Malaysia's largest trading partner, as a rival: "I think the reality of it all is, there is enough growth that will go around. So, everybody will get some pickups on that one." He added that Malaysia is not going "head on to the tail end of the high-end competition, where maybe the U.S. is bringing all the parts."

A single town, and a single material

Capacity also depends on materials that rarely make the headlines. Spruce Pine, North Carolina, is home to high-purity quartz mines that produce ultra-pure quartz, used to make the crucibles that hold molten silicon as it is turned into ingots. Tom's Hardware reported on 1 October 2024 that Hurricane Helene brought over two feet of rain to Spruce Pine, cutting off roads and leaving residents without power. It was difficult to assess damage to the two mining companies, Sibelco North America and The Quartz Corp. The Quartz Corp told The Verge it was too early to determine whether its high-purity quartz production facility had been affected, according to Tom's Hardware, which added that Sibelco had purportedly confirmed a suspension of operations. Tom's Hardware also noted a 2008 fire in the town disrupted the flow of high-purity quartz to the world market, and that a production halt today could have more severe consequences as the world needs more chips than in 2008.

Construction Physics examined the claim that all semiconductor manufacturing depends on Spruce Pine quartz, in a piece published on 8 November 2024. Its verdict: Spruce Pine quartz is not quite an irreplaceable linchpin, but alternatives are all some combination of not yet developed, not quite as good, and not quite as cheap. Cutting off the supply probably would not choke off semiconductors completely, but it would mean yields going down and costs going up, according to that analysis.

The same piece lays out the process constraints in detail. Most modern semiconductors are made from silicon that is 99.9999999% pure, known as 9N, for microchips, and somewhat less pure, 6N to 8N, for solar cells. Monocrystalline silicon is almost entirely made by the Czochralski process, in which polysilicon is melted in a quartz crucible at more than 1400 degrees Celsius. Quartz crucibles cost several thousand dollars apiece and are limited to around 400 hours of operation. One paper estimated that 30% of the manufacturing cost of monocrystalline silicon ingots came from the crucibles, Construction Physics reports.

Materials, not just machines

Chipmakers are also expanding capacity in specific materials. Texas Instruments announced on 24 October 2024 that it had begun production of gallium nitride power semiconductors at its factory in Aizu, Japan. Together with its existing GaN manufacturing in Dallas, Texas, that will see TI internally manufacture four times more GaN-based power semiconductors as Aizu ramps to production. Mohammad Yunus, TI's senior vice president of Technology and Manufacturing, said in the announcement that the company had "successfully qualified our 200mm GaN technology" and aims to grow internal manufacturing to more than 95% by 2030.

TI said it had piloted GaN manufacturing processes on 300mm wafers earlier in 2024, and that its expanded GaN processes are fully transferable to 300mm technology. The company also said it has committed to use 100% renewable electricity in its US operations by 2027 and worldwide by 2030. None of this changes the arithmetic of a fab, but it changes who can supply it.

Put together, the record suggests capacity is best understood as a stack: the process line and its layer times, the national incentive programs paying for new plants, and the mines and crucibles feeding them. The IBM example sets the historical benchmark. Modern fabs, by IEEE Spectrum's own numbers, are not yet beating it.