The UK government may be onto something with its strategy to support the domestic semiconductor industry. In a strange twist of fate, some experts are starting to say that its approach makes sense.

Britain's long-awaited semiconductor strategy was published in May, promising funding and a focus on areas where the country is seen to have strengths in the chip business, rather than pouring money into building massive fabrication plants and tempting industry giants like TSMC and Intel to set up shop locally.

Reception was lukewarm, with many criticizing the £1 billion ($1.2 billion) funding as being too small compared with the sums being coughed up by the US and EU programs, while others just welcomed the fact that the government was actually doing something.

However, in a recent article for Nikkei Asia, two technology think tank executives, Jan-Peter Kleinhans and Pranay Kotasthane, asserted that massive chipmaking subsidies are not the answer to supply chain worries, and that the UK might actually be taking the right approach.

The article claims that the subsidized back-end investments so far announced are "unlikely to shift the needle on a global scale," with much of the chip output likely to still come from Taiwan and other parts of Asia. It also points out that chips made in the US by TSMC are likely to be 30 percent more expensive than those made in Taiwan, and the final assembly and packaging is often done in Asia as well.

"Governments have chosen a costly path with little consideration for second-order effects on consumers and businesses," the authors conclude. "The UK's strategy of focusing narrowly on certain segments of the supply chain where it has a comparative advantage offers an alternative approach."

Charles Sturman, CEO of industry association TechWorks and chair of the Semiconductor Leadership Group, agrees that the UK government is "moving in the right direction now," although he does not believe they have got everything right.

"The government knew they needed to have a sector strategy for semiconductor," Sturman told The Register. "But they're politicians and civil servants, and even someone who works in the chip industry won't understand all aspects of it; if you're a design guy, you don't know about fabrication processes, and vice versa. It is highly complex.

"What I would have preferred the government to do would be to engage in some kind of industry consultation process. But, in fact, that's what they're doing now," Sturman said, referring to the Semiconductor Advisory Panel that met for the first time in August.

"They have this strand of industry consultation around semiconductor manufacturing now, so I think we're in a better shape than we were perhaps when they first published the strategy," he added.

Another of the areas the UK strategy identified is helping chip startups to get their products over the finishing line to market, which is now being addressed to an extent via the ChipStart program.

"There's a lot of risk, and it can take a long time, maybe five to ten years," Sturman said. "And to design a chip, the EDA (electronic design automation) tools are stupidly expensive. If I go to a foundry to make a test chip, that's also really expensive."

To help here, ChipStart is run by SiliconCatalyst, which has already done deals with the EDA companies, IP companies, and foundries to offer low-cost access to those capabilities, so that startups on the program can at least get to their first prototype, he explained.

Advanced packaging is another area government identified as an opportunity for the UK.

"Advanced packaging is one of those massive inflections because instead of trying to build bigger and bigger chips on one piece of silicon, the industry has recognized that you can cut those chips up into smaller functional blocks," Sturman explained. This is the so-called chiplet approach, which involves assembling those pieces on a substrate, almost like a circuit board inside a chip package.

There are other areas of the semiconductor industry where there is high value and moderate volume to be made, according to Sturman, areas that he said the giants like TSMC don't touch, such as silicon prototyping, photonics, MEMS (micro-electromechanical systems), quantum chips, and silicon carbide (used for power electronics in automotive).

This is where compound semiconductors comes in, another subject of government focus, and which Sturman said is important because photonics, MEMS quantum, and silicon carbide are all based on compound semiconductor materials.

One criticism Sturman raised of the government's semiconductor strategy is that it appears focused on early stage innovation. "We need to look at the whole ecosystem in terms of SMEs and other companies in the UK that are struggling to expand or to upgrade their facilities, or to compete on a global scale, because globally, it's not a level playing field," he said.

"Most developed nations have got subsidies. Whether it's tax breaks or land tax breaks or grants, there are subsidies to the tune of 20 to 50 percent if somebody wants to build a fab, or create a chip design company, and that makes it quite difficult for a company in the UK to compete and scale up, which is a real concern."

On the matter of having a supporting semiconductor ecosystem, Sturman said that startups need not just partners and suppliers, but also capital and finance to support them through that scale-up phase.

"Early stage seed funding startups from universities, we're quite good at that already, we've got a lot of that going on. Where I see the issues are Series B. if you're a chip company trying to raise Series B funding, to scale up into mass manufacturing for global competition, you've typically got to go to the US to get that money, you're never going to get that in the UK today," he said.

If Britian does not get that in place "then sooner or later, the companies end up getting bought by a foreign interest. And then a few years down the line from there, there won't be any staff left. So I think it's important in all of this, to think about the longer term," he said.

Should the UK invest in fabrication plants then? Sturman said there was every reason to, but there was no point in trying to outdo Taiwan or Korea in pouring money into costly cutting-edge facilities.

"We're never going to spend $10 billion on building a state-of-the-art 2nm fab for making chips to go into iPhones. We could build one with a moderate technology node like 28 to 40nm. That would satisfy the needs of the bulk of what goes on in the UK today, which is mixed-signal, analog and digital communications," he said.

There's a very lucrative industry around the world in semiconductor chips, "which is not the super high volume commodity stuff," he added.

Overall, Sturman seems to believe that what the Semiconductor Leadership Group would like to see is broadly in line with government thinking, but that it is "about capturing more of the semiconductor value in the economy in the UK," he said.

That's not just about revenue, but having more impact on the UK economy for supply chain resilience and national security. That can be difficult because there are so many different components to the supply chain in things like cars or aircraft, but "the more you have, the more bargaining chips you've got if you have to go to another country and say, hey, we need some more supply of what you've got."

It is also worth remembering what Gartner's former vice president for semiconductors and electronics, Richard Gordon, told us in frustration after reading the parliamentary committee report on the state of the UK semiconductor industry.

"It's like they've just discovered that semiconductors are important," he said, adding that "successive governments have failed here, and they haven't learned anything over the last 30 years or so ... This report seems to be saying that these are our only choices now, so let's do that." ®

https://www.theregister.com//2023/10/31/uk_semiconductor_strategy/