JAMES ASHTON: Can we cash in on microchips? UK is right to focus on its strengths such as design and materials, but it must afford the industry time
- Microchips have taken over the world
- In 2021, supposedly riven with shortages, more than 1tn chips were installed
- The most advanced chips are required to power artificial intelligence
A billion pounds doesn’t get you very far in the microchip industry. The mighty Intel spends that much every 26 days on research and development alone.
By setting aside so little over the next decade to fund its semiconductor strategy, the UK government will hardly trouble American and Chinese political strategists sparring for global supremacy in this space.
Microchips have taken over the world. Even in 2021, the year supposedly riven with shortages which clobbered the production of cars and computer games consoles, more than one trillion chips were installed in devices to process and store information.
And there is more to come, as the most advanced chips are required to power artificial intelligence (AI) that churns through data to make lightning-quick decisions.
Looking ahead: The most advanced chips are required to power artificial intelligence that churns through data to make lightning-quick decisions
It was the pandemic-induced production blip that persuaded politicians to invest in the vital staple of the digital economy.
The UK can’t hope to compete when the best chip plants cost £16billion to build. And even if it found the cash, there will be a glut of factories by 2030 – perhaps too many.
But money is not everything. If it was, Arm, the Cambridge-based company whose chip designs are used 1,000 times a second, would never have prospered.
Born out of the ashes of Acorn Computers, its only political leg-up of sorts was that Acorn’s BBC Micro sold strongly when the Thatcher government chose it as one of two homegrown devices to qualify for a 50 per cent government subsidy with the aim of installing a computer in every secondary school by the end of 1982.
Acorn at least had a little cash to invest in a processor design it thought would propel it forward. But precisely because it didn’t have too much money, that design was created to be low-powered and low-cost, following the trend for a simple, reduced instruction set computer (RISC).
The Acorn RISC Machine (Arm) project would likely have been wound down if Apple hadn’t been looking for a processor design for a new portable device. Instead, it was spun out into a new company in November 1990.
A decade later, political interest was stirred when Stephen Byers, the then trade and industry secretary, opened the company’s new headquarters. But by that point the company’s shares had entered the FTSE 100, and its product featured in Nokia mobile phones.
Arm, now owned by SoftBank of Japan, has chosen to list its shares in New York, not London. To create another UK semiconductor champion, it is important to understand that this complicated, capital-intensive industry is built on tiny specialisms.
And the UK can learn from Japan which maintains a vital role in the supply chain, producing hydrogen fluoride – used when the chip design is etched on a sliver of silicon – and the light-sensitive to photo-resistors that project the intricate circuit patterns.
The UK is right to focus on its strengths such as design and materials. What it must afford the industry is time. If poverty again breeds creativity, the least the UK can do is buy British. Another promising company in this field, Graphcore, hopes the government will do so when it comes to equipping its new supercomputer project.
After all, the original US chip industry only gained traction when President John F. Kennedy invested heavily to beat the Soviet Union in the space race. More than 60 years on, the race continues.
- James Ashton’s book, The Everything Blueprint, the story of Arm, is published by Hodder & Stoughton.
