eeNews Europe talks to Americo Lemos, newly appointed CEO of wafer maker IQE in Wales.
Lemos has been in post for six months having joined IQE from Globalfoundries in January 2022. Lemos had previously worked for Qualcomm, Intel and Texas Instruments. He is an executive with much experience of the fast-moving semiconductor industry that lies immediately downstream from IQE in the supply chain.
To begin with we pointed out that an operational profit for IQE in 2021 had been nullified by exceptional items to do with discontinued or shelved projects and site relocations. We asked if such measures are all behind the company now. “We are in business to make profits and we’ll do that two ways: rationalization of our footprint and scaling our footprint.” Lemos said.
Lemos explained that under IQE’s previous CEO, Drew Nelson, the company had grown steadily by a series of mergers and acquisitions which brought with them multiple sites. And under Nelson’s leadership a couple of decisions had been made to rationalize. These are: the closure of a site in Singapore to focus IQE’s Asian footprint in Taiwan; and a plan for the consolidation of MBE [molecular beam epitaxy] production at IQE’s North Carolina site alongside the closure of its Pennsylvania site in 2024.
“I revisited the [rationalization] decisions when I joined and agreed they are right. Can I rationalize further? Maybe, but I think we are pretty much there. We now need to scale. And we have [at Newport] a modern-built site, we have the space, the power budget and the infrastructure in place.”
During our visit to the Newport site we had seen the first 10 metal-organic chemical vapor deposition (MOCVD) machines producing epitaxial wafers but the potential size of the cleanroom is enormous. The cleanroom has been configured for just 20 machines right now but could be expanded easily to accommodate 100 machines. And that is on one floor of a two storey building.
And Lemos is confident that the compound semiconductor market is coming towards IQE and that will allow that expansion to happen over the next three-to-five years.
Could IQE continue to grow by M&A?
“We could. There are lots of markets where we are not playing, where compound semiconductors are starting to have an impact. We need to get into automotive and microLED markets, partly to diversify and protect us from particular market swings.” Lemos is aware that a slowdown is happening in the consumer and smartphone market where there is a lot of unit demand for compound semiconductors.
“There is also a captive market in epitaxy. I lot of device companies have their own MOCVD machines. It is up to IQE to step up. There is a huge opportunity there,” said Lemos.
But what about the global signs of recession and stress that we are starting to see both in the general economy and in the semiconductor sector? Is now the time to be putting in semiconductor manufacturing equipment for economies of scale that may then lie idle?
“We know there is going to be the impact of global inflation. We see the softening of smartphone demand and slowing due to Covid lockdowns in China. But the megatrends will continue: the move to ‘net zero’ carbon drives energy efficiency; intelligent transport and the metaverse drives efficient power and high frequency communications. This may slow down but it won’t go into reverse.” And Lemos said the trends all favor compound semiconductors, for which IQE is the only global player making the wafers.
Lemos acknowledged that IQE needs to keep a careful watch ahead by engaging with customers. That’s partly why IQE entered into a long-term relationship with photonics company Lumentum Inc., he said. IQE is supplying Lumentum with wafers for 3D sensing, automotive lidar, and optical networking.
IQE also made a deal with Lemos’ previous employer Globalfoundries. This deal will allow GaN-on-Si to be offered at Globalfoundries’ wafer fab in Burlington, Vermont, using wafers supplied by IQE.
“They give us more visibility into where we may experience headwinds. We need to be diligent across the entire ecosystem of how products get made. But a slowdown may accelerate the shift from silicon to compound semiconductors,” Lemos concluded.
Finally, we asked about the possibility of a compound semiconductor foundry in south Wales. An open-access pilot foundry and had been part of the plan at Newport Wafer Fab but appears to have gone on to a back burner since that company was taken over by Nexperia BV in July 2021.
Lemos said: “It is important to build a complete ecosystem like Taiwan and Israel have done. IQE is an excellent company but it is not an ecosystem. In Wales there’s Cardiff and Swansea Universities then there is the Compound Semiconductor Cluster but that is not enough. IQE’s wafers [from South Wales] are 100 percent exported. My proposition to the UK government is that we need to build a foundry for compound semiconductors in Wales – a pure-play foundry that can process IQE wafers.”
Lemos said that it could be wholly- or partly-owned by IQE but it would require public-private partnership to get such a project started. “We can put a pure-play foundry into our building. We have the space,” said Lemos.
“Alternatively, we can build a ‘greenfield’ fab,” he added, while pointing out that such a choice would add about two years of delay for the planning and regulation and for the construction of the shell building.
If IQE went on its own, or was the majority shareholder in such a venture, would it be competing with its customers? “We wouldn’t be competing with customers. Compound semiconductor foundry Win [Semiconductor Corp. in Taiwan] is doing its own epitaxy.”
Compound semiconductor devices can be highly complex, multi-layered structures requiring close cooperation between epitaxial wafer maker, foundry and chip company. IQE would be enabling downstream customers, Lemos said.
Lemos pointed out that the UK is a major country in Europe that has yet to make a move towards securing its technology future. And Wales represents an opportunity to invest in compound semiconductor technology whose time is coming and at a lot less cost. “It is a lot less expensive than building silicon infrastructure,” Lemos added.
So how much does Lemos think should be spent: a few tens of millions of pounds or a few hundred millions of pounds for R&D and pilot production?
Lemos answered: “A high-level figure would be a billion pounds. There is no point in building sub-scale. We want to create the TSMC of the compound semiconductor business,” He added: “The obvious place to do it is South Wales.”
Typical levels of government subsidy in major semiconductor capital expenditure projects have run at 30 percent and up to 40 percent. It may be obvious to Lemos that this would be investors’ and tax payers’ money well-spent but the UK government is in transition of Prime Ministers. This could disrupt any discussions that IQE has managed to get going. The UK is also under a right-of-centre of government that is generally averse to large subsidies for industrial capital expenditure.
“It is an opportunity for the new Prime Minister to have quick win,” Lemos said, adding that according to the Semiconductor Industry Association for every direct employee in a wafer fab between 5 and 6 indirect jobs are created.
Lemos acknowledge that compound semiconductor wafers would still have to leave the region for outsource semiconductor assembly and test (OSAT) which is not done in volume in Europe. Lemos said that when building an ecosystem it is necessary to proceed one step at a time.
But Lemos said it is also necessary to proceed to quickly: “The market doesn’t wait for us. This is something we would need to get going in the next 12 months. But this is a tremendous opportunity. We can leverage existing infrastructure in a built-for-purpose building. It makes sense do it this way.”
The technical storage or access is strictly necessary for the legitimate purpose of enabling the use of a specific service explicitly requested by the subscriber or user, or for the sole purpose of carrying out the transmission of a communication over an electronic communications network.
The technical storage or access is necessary for the legitimate purpose of storing preferences that are not requested by the subscriber or user.
The technical storage or access that is used exclusively for statistical purposes.The technical storage or access that is used exclusively for anonymous statistical purposes. Without a subpoena, voluntary compliance on the part of your Internet Service Provider, or additional records from a third party, information stored or retrieved for this purpose alone cannot usually be used to identify you.
The technical storage or access is required to create user profiles to send advertising, or to track the user on a website or across several websites for similar marketing purposes.