Sorry Brainport, K&S momentarily has its sights fixed on Silicon Valley

The relocation of Kulicke & Soffa’s production to Singapore painfully exposes the weakness of our high-tech industry and the European chip strategy. European parts, manufacturing capacity and engineering hours are simply too expensive for a back-end player in a market dominated by the front-end, I wrote recently.

K&S was trying, though. It put two heavy projects on the shoulders of its Eindhoven engineers. For a major end customer (in all likelihood Apple), it was developing a new placement machine for backlighting. In addition, it needed a new version of its iFlex pick-and-place machine. Two massive projects, which turned out to be millstones. As a ‘reward,’ production of the iFlex is moving from Eindhoven to Singapore.

From this, it’s easy to conclude that Europe can’t compete in a cutthroat part of the back-end market – assembling printed circuit boards. However, in a world where the international order is being shaken up and countries and continents are striving for sovereignty, other factors come into play. When it comes to R&D investment and market opportunities, for example, the US is beckoning.

K&S is already working with UCLA on advanced thermal compression bonding (TCB) machines. The millions from the US Chips Act provide an additional reason to focus on America’s West Coast. As recently as last July, the US awarded 1.6 billion dollars in R&D grants to boost industrial activity for advanced back-end technology. Among the beneficiaries is the world’s largest packaging and testing company, Taiwan’s Advanced Semiconductor Engineering. With the money, ASE is doubling its capacity in California to help customers such as AMD, Apple and Nvidia. Competitor Amkor previously received 400 million dollars for its new plant in Arizona.

Stimulated by the US Chips Act, K&S is also now joining Japan’s Resonac, a new consortium of ten industrial partners for R&D in advanced packaging in Silicon Valley. This group of machine builders and material and tool suppliers includes KLA. The consortium partners are setting up an R&D center in Union City, California, to be launched in 2025. Although the amount of the investment is undisclosed, we do know the typical sum the National Advanced Packaging Manufacturing Program (NAPMP) makes available for US projects to boost local R&D in advanced packaging technology: 150 million dollars. In Europe, we spend that kind of money on front-end technology; the back-end gets ten times less.

The decision of K&S to take a hard look at its Eindhoven operations should also – and perhaps especially – be viewed from the paranoid perspective in which the back-end has to operate. The world of chip packaging is what business gurus call “red ocean,” a market that’s colored with blood due to fierce competition and low entry barriers. The fish in this sea are diligently searching for a blue ocean, where they can grow with innovative new products, out of reach of voracious competitors.

On the back-end, however, rests a curse. Once a promising niche is found, disaster tends to strike. Competitors follow the pioneers and the blue ocean soon turns red. In TCB, K&S is currently quite successful. The innovation is focused on a market where many back-end companies are seeking business: packaging AI chips. In today’s chip market, that’s about the only place where end customers like Amazon, Facebook, Google and Microsoft are still investing ruthlessly.

In emerging markets like these, back-end equipment manufacturers must make ruthless choices. The high-end segment is operating globally and has to maintain an installed base of thousands of machines, spanning many generations, with dozens and sometimes hundreds of product lines. So packaging machine builders have to weigh their investments. A gamble on a blue ocean can also be a dead end, or mistimed. K&S tried it with placement machines for mini and microLEDs, but that market turned out to be too small – or they simply were too early, who’s to say.

Either way, for K&S, the development of TCB bonders takes priority right now. That’s where the high margins are. The company is currently one of the frontrunners in TCB, and it looks like it’s doing better than Besi in terms of timing and customer acceptance. The competitor from Duiven is aiming at another blue ocean, hybrid bonding, a prohibitively expensive technology by back-end standards that’s also not yet mature.

For those engineers among our readership who still think the Dutch can beat the Americans in placement technology and mechatronics: as far as the back-end industry is concerned, I’m afraid that we should harbor no illusions.

Main image credit: Kulicke & Soffa