Chips with everything — but UK manufacturers are scarce

In the 1980s, when Arm, the UK’s flagship chip architect, was a titch called Acorn Computers, a car might contain three chips. Today it has 2,000-plus and when semiconductor supplies ran short during the pandemic, the car industry stalled.

The demand for ever-smaller, evermore powerful microchips is unstoppable whether economies tank or banks melt into crisis. “Chips are more important than oil,” says Ian Lankshear, chief executive of Aim-quoted chip designer EnSilica.

Consultant McKinsey predicts chip sales will reach $1tn by 2030, driven by three sectors: automotive, computation and data storage, and wireless.

Even eight years ago, when Arm was valued at about £15bn, I only half believed Simon Segars, Arm’s then chief executive, when he talked of the “internet of things” and the demand for smart toasters and even smarter phones. But I did believe in Arm’s strategy of charging customers royalties and licensing fees for must-have processor designs.

Now politicians lament that Britain will lose its only big semiconductor success when Arm relists in the US at a mooted value of $40bn. They fret the UK will be forever squeezed between US device makers and Asia’s manufacturing behemoths and dependent for key strategic components on a global chain buffeted by geopolitics and geoeconomics. Tellingly, Joe Biden recently provided $39bn in funding for US semiconductor manufacturing and $24bn worth of manufacturing tax credits via the Chips and Science Act.

Still, a report by Geoffrey Owen (former FT editor) for the Policy Exchange says the UK has built niche strengths and competitive advantage in designing compound semiconductors for specialist applications in, for example, defence and telecoms.

That made me wonder whether Acorn-sized chip designers are lurking unnoticed on the UK market.

Not many, says Bob Liao, analyst at brokers Zeus Capital, gloomily. The UK will always struggle against overseas enterprises backed by indulgent governments and deep-pocketed financiers.

Tech start-ups have it tough in Britain, agrees Graham Curren, chief executive of Sondrel, designer of application-specific integrated circuits (Asics). The orchestrated rescue of the UK arm of Silicon Valley Bank only highlights the gap in funding for Britain’s high-tech hopefuls.

The problem, tech entrepreneurs complain, is UK investors over-focus on returns on capital, are suspicious of cash-hungry companies jostling for a footing in a competitive industry and sell out too quickly.

Readers may recall Dialog Semiconductor, which was flogged to Japanese rivals in 2021. In 2016, Arm was sold to SoftBank of Japan for $32bn, or about 24 times earnings. Wolfson Microelectronics was picked off by a US competitor for about £300mn in 2014, the same year that Qualcomm scooped up Cambridge Silicon Radio for $2.4bn.

The problem, tech entrepreneurs complain, is UK investors over-focus on returns on capital, are suspicious of cash-hungry companies jostling for a footing in a competitive industry and sell out too quickly

And then there was Imagination Technologies, once one of the UK’s biggest listed tech companies, which crumbled when Apple stopped buying its IP. It was sold to venture capitalists in 2017 for about £550mn. 

Investors will counter that, all too often, chip designers struggle to tot up their revenues from partnerships and complex multiyear licensing and service contracts.

Arm-wannabe Alphawave IP, the Toronto chip creator which floated shares in London at 410p in 2021, has not seen its stock recover since FT Alphaville questioned sales linked to related parties. Its shares are about 120p. 

Patchy earnings and multiple cash calls have persistently tarnished Cardiff-based IQE, which manufactures wafers for fibre optic sensors. The shares at 25.85p are far off 2018 highs of 170p.

That said, Sondrel and EnSilica, 20-year old semiconductor veterans that have both made junior market debuts in the past year, are intriguing.

Berkshire-based Sondrel works with clients in futuristic sectors such as automated cars and artificial intelligence. It ran into losses during the pandemic and now wants to break into overseeing production of its semiconductor designs. Growth is constrained by headcount, says Curren.

Seeing a chip through to manufacture leverages the skills of engineers, potentially magnifying revenues at little extra cost. A client will spend up to $30mn for a design and perhaps $100mn to take it into production.

EnSilica, like Sondrel, is an Asic designer of chips mixing digital and analogue signals — for use in cars, broadband satellite communications, insulin pumps and heart monitors.

Both Lankshear and Curren see opportunities as European and US customers take control of supply chains, bringing production closer to home.

Broker Allenby Capital forecasts EnSilica will earn £500,000 pre-tax by 2024 on sales of about £23mn, against £15mn in 2022.

Broker Cenkos reckons Sondrel will make £2.4mn pre-tax profits in 2024 on revenues of about £40mn up from £17.5mn in 2022. Sales could double again by 2025.

I don’t downplay the challenges. One of the biggest is recruiting talent. Chip engineers are scarce and both Sondrel and EnSilica have ranged far to secure scarce skills.

Keep in mind CML Microsystems, another mixed-signal chip designer that has recently switched to Aim after years languishing on the main market.

It has cash, pays a dividend but still only raked in £17mn in sales in 2022, making £2mn in adjusted pre-tax profits. Its shares, valued at close to £90mn, have just returned to 2000 levels.

I accept that investing in smallish chip companies is a leap of faith. While oaks do grow indeed from acorns, they do so slowly, and none may grow as mighty as Arm.