UK government is in semiconductor mess of its own making
Newport Wafer Fab makes for an uninspiring name for a business and an unlikely subject for a totemic battle over national security and strategic economic interests.
But that is where the company, which unsurprisingly fabricates wafers in Newport, finds itself.
Business secretary Kwasi Kwarteng has until the end of June to use new national security powers to reverse the Welsh plant’s purchase by Nexperia, a Dutch subsidiary of Chinese company Wingtech. At the same time, another department is working on a national strategy for semiconductors, amid a global shortage and mounting concerns about security of supply.
The UK government’s new oversight of dealmaking was never going to be able to steer clear of questions regarding industrial policy. But they might have hoped to avoid a direct collision quite so quickly.
The crux of the Newport case is that the plant appears pretty irrelevant on a narrow view of national security and yet significant through the wider lens of economic strategy in a high-tech sector the government says it wants to develop.
That’s a problem given it has been at pains to separate the two. Deal-vetting powers are often used overseas to protect industrial policy priorities. But economic considerations were removed from early guidance by the time the National Security and Investment Act powers came into force this year.
The facility, built by government-backed Inmos before being privatised in the 1980s, makes legacy silicon wafers that are shipped to Asia to be assembled into chips. This is decidedly lower-value stuff, far from the cutting edge of manufacturing that predominantly happens in Taiwan. Even some within the sector concur with Stephen Lovegrove, the government’s national security adviser, in seeing no direct threat from its sale.
But a post-pandemic shortage of semiconductors and heightened geopolitical concerns have changed the tenor of the global conversation. The mood is such that nine US congressmen have raised the Newport sale in a letter to President Joe Biden. “The US and EU are talking in terms of strategic economic significance and pursuing ambitious strategies to retain and reshore semiconductor capacity,” says George Dibb, head of the Centre for Economic Justice at the IPPR. “This situation shows the weakness of the UK’s tool kit.”
The UK isn’t going to be able to spend its way to semiconductor self-sufficiency: the sums are too big and the sector too complex. Its strength lies in cutting-edge chip design (hence the angst over what happens to Arm). Then there is the opportunity to nurture developing specialisms in new compound semiconductors, which combine two or more elements and can outperform silicon in some applications when it comes to power, light and speed, including in the automotive industry.
The hope was that Newport Wafer Fab would plug into those efforts, as part of a cluster of businesses in south Wales. We know that the government — and Kwarteng — likes clusters, because they told us so repeatedly in last year’s innovation strategy: the Welsh one, which comes with a public-funded compound semiconductor applications catapult, even got a mention.
Nexperia, in fairness, says it only pulled out of one project, involving potentially sensitive technology; it has invested £80mn and added jobs. But the group is fundamentally focused on churning out commodity chips for its clients, not a so-called “open access” model that would contribute to the future thriving of the industry in south Wales: it will honour existing contracts but not commit to extensions or new research activities.
Everyone agrees that a volume manufacturer is important to give the cluster heft (and the Newport plant has struggled over the years). But the idea was really that those developing new technology, including a company founded by the former owner of the Newport fab who has an option to use part of the facility, would be able to test and scale up production in one place. Could that happen in other locations? Perhaps. But the point of clusters is proximity. And why spend time and potentially public money recreating capacity elsewhere?
The government seems to want to put clear water between the deal and a semiconductor strategy, in the pretence that these are unconnected issues. They are not. After all, what is the point of strategies to develop innovative industries if the government lacks the tools or inclination to then safeguard them?