Customers before capacity: what a UK chip fab can teach distributors

Steve McCue has been chief financial officer of Pragmatic Semiconductor since March 2025, and his job is the hardest one in hardware: making years of research pay. In an interview with Raconteur published on 24 September, he described a company still spending on the next two generations of its chip while it increases production of the current one. Pragmatic, he told Raconteur, is “developing technology that changes where semiconductors can be used”.

The technology is the FlexIC, an ultra-thin, flexible integrated circuit. It is made at Pragmatic Park, a 60,000 sq m site near Durham that opened on 27 March 2024 as the UK’s first 300mm wafer fab, with room, the opening release says, for up to eight further production lines. Pragmatic’s Series D round, announced on 6 December 2023, raised £182m ($231m), led by M&G’s Catalyst fund and the UK Infrastructure Bank, since renamed the National Wealth Fund, to build the third and fourth lines. The release calls it the largest venture round ever raised by a European semiconductor company.

McCue told Raconteur that Pragmatic spent about 15 years on research before commercial production, and that his task is to keep funding the next generations – Gen 3 is in the market, Gen 4 is due in 2026 and Gen 5 is planned for 2027 – while managing cash from day to day. A capital-intensive business, in his account, needs a Plan A, a Plan B and a Plan C. Some fab equipment, he told Raconteur, has lead times of around 15 months, so the investment decision is made long before the asset is installed, and any forecast will have moved by then. He ran manufacturing operations in Scotland and Singapore before his CFO roles, he told Raconteur, and argues that finance chiefs who have carried a profit and loss account bring a perspective finance-only careers do not.

The line channel leaders should pull out is a different one. McCue told Raconteur that a deep-tech company should have its global customer relationships in place, and its strategic customers aligned, before it expands aggressively. Customers before capacity. It is the rule Pragmatic is applying to a fab, and it is the rule a distributor should apply before it stocks anything that depends on new silicon.

The sovereign-hardware story has a timing problem

The government’s National Semiconductor Strategy, published on 19 May 2023, commits up to £200m over 2023–25 and up to £1bn over the following decade, and cites Pragmatic’s $125m Series C, completed in December 2022, as the kind of scale-up it wants more of. Every UK integrator and distributor has heard the pitch: domestic chips, resilient supply, a route that does not run through Taiwan. Pragmatic is the best UK example of it.

Look at the gap between capacity and orders. The Gen 3 launch release of 4 March 2025 puts the Durham foundry’s capacity at billions of chips a year, the company’s own figure. The first FlexIC-based radio-frequency identification (RFID) product lines, from the French inlay maker Tageos, were announced on 15 April 2026, with volume orders expected from Q3 2026, roughly two and a half years after the fab opened. None of that is a criticism of Pragmatic; it is what McCue means by lining the customers up first. It is the timeline to keep in mind when a vendor says a new category is about to take off.

Scaling the fab has not been smooth either. On 8 September Pragmatic appointed Dr Paul James, formerly managing director of Vishay Newport, as executive vice-president of manufacturing and operations, replacing Shane Geary after nearly three years overseeing the fab’s opening and ramp, the release says. The person running the ramp changed just over a fortnight before McCue’s interview appeared.

What the rule means for a distributor

Distributors take on hardware lines the way chip companies build fabs: on a forecast, with cash committed before a unit sells. The difference is that a distributor rarely writes down its Plan B. It signs, takes the opening order and learns a year later whether the vendor’s pipeline was real. AI-capable edge devices, private 5G kit, smart labels, satellite terminals: each arrives with a deck that shows capacity and calls it demand.

McCue’s discipline translates directly. Before Pragmatic builds its next line it wants to know who is buying from the lines it has. Before a distributor stocks a new silicon-dependent category it should know which of its resellers will buy it and in what volume. If that list is short, the category is a vendor’s hope, not yet a distributor’s business.

The checklist before you take the line

• Named demand, not addressable market. Ask the vendor for its committed customers by name and volume, then map them to your reseller base. A market-size slide means the vendor has capacity and wants you to find the customers.
• The half-forecast test. Model the line at half the vendor’s number and ask whether the working capital still makes sense. That is Plan B. Plan C is what happens to the inventory if the category never turns: return rights, price protection and a buy-back clause, agreed in writing.
• One source or two. McCue told Raconteur that Pragmatic manufactures at a single UK site. Every hardware line has one. Ask where the second source is, and how long a re-spin takes if the first one slips.
• Who owns the ramp. A change of manufacturing leadership at the vendor, as at Pragmatic in September, is a conversation rather than a panic. Ask what changes in the delivery plan.
• The exit before the entry. Agree how the line ends, at what notice and who holds the stock, before the launch, not after the first write-down.

The National Semiconductor Strategy will keep telling the channel to care about sovereign hardware, and it should. A fab, though, is a promise of capacity, and capacity has never paid a distributor’s bills. Customers do. Line them up first, then order the stock.