With the loss of much of its manufacturing base to Asia, Europe faces an uncertain future in semiconductors.
Does the European semiconductor industry have a future, or is it in an inexorable slide into irrelevance? It's a tough question, but important if you cleave to the modern notion that 'we're all knowledge workers now'. If Europe cannot create sustainable businesses in such a knowledge-intensive sector as semiconductors, what prospects then for a European knowledge economy?
Before we go too far, it pays to define the 'European semiconductor industry' we're talking about here as indigenous companies that were started in Europe and that remain headquartered here. The mid-term test of the viability of the European chip industry has to be whether it can sustain a critical mass of manufacturing, markets, skills, financiers and entrepreneurs to enable companies started here to grow into thriving listed global businesses.
Speaking at the launch of the European Institute of Innovation and Technology (EIT) in Barcelona at the end of May, EIT governing board member Professor Manuel Castells pointed out that Europe's share of global gross domestic product has declined from 29.9 per cent in 1980 to 21.4 per cent now. Meanwhile, the developing world's share has risen from 7.1 per cent to 22 per cent over the same period.
The same shift of global focus is reflected in the semiconductor industry. Speaking at the International Semiconductor Forum, organised by the GSA and the IET in mid-May, Heinz Kundert, president of SEMI Europe, which represents the equipment makers, said: 'Over the last 10 years we can clearly see the industry's decline. Ten years ago about 20 per cent of the device and the equipment market was based in Europe. Now it's 10 per cent of the equipment market and less than 10 per cent of the device market.'
David Baillie, CEO of CamSemi and chair of the GSA European leadership council, who chaired a debate on this topic at the forum, said: 'We have been, and in some areas still are, leaders. The question is, where do we go from here?
'There's a shift of power relating to resources and where technology is being developed, manufactured and exploited,' he added. 'So how do we foster innovation and convert that into world-class businesses?'
Stan Boland, president and CEO of wireless baseband chip company Icera, said that the future growth of European industry is hobbled by the fact that Europe's venture capitalists only have around 10 per cent of the funds of their US competitors, and that the cost of starting a global semiconductor business has risen tenfold in ten years, to around $200m.
'The opportunity is much, much worse now than 10 years ago,' he said. 'We're also squeezed between continents, where Asia has a bias to low costs and low average selling prices (ASPs), while in the US the established players are very serious about keeping their market positions. They aim to bully or litigate their way to market dominance.'
As Europe is neither low cost nor litigious, Boland argues that we have to find another way forward. He says Europe is good at creative engineering, developing novel chip architectures and protecting them well, which enables a balance of low ASPs and good margins. Europe is also close to sources of finance, and has governments that can support the industry through tools such as R&D tax credits.
He proposes two ways forward for European industry. The first is a 'coral reef' approach where there are many high-quality analogue, radio frequency, and power companies producing parts with high margins but not necessarily for large markets. The second is to go for 'deep sea' targets, trying to build aggressive future public companies by addressing mass markets with strongly differentiated products, particularly for key European specialities such as the telecoms and automotive markets.
Jalal Bagherli, CEO of Dialog Semiconductor, agreed that Europe is rich in systems knowledge: 'We can play to European strengths in mixed signal, wireless, automotive, industrial and eco-friendly markets.'
But he was quick to point that 'coral reef' companies still need global ambition: 'You can dominate in the niches you choose to play in at the world scale. But you can't just be an excellent company in the UK, France or Germany - you have to compete globally.'
Europe's systems knowledge is in part based on its rich manufacturing heritage, and in part based on a social outlook that values things such as universal healthcare and public transport. As Europe's semiconductor manufacturing base has declined, Belgian nanoelectronics research institute IMEC, among others, has promoted the idea of 'More than Moore' as a way forward for companies that can no longer afford 'More Moore', the relentless pursuit of smaller device dimensions in bulk CMOS as per Intel founder Gordon Moore's prediction.
More than Moore suggests ways to add value by building new functions, such as biosensors, on top of mainstream and even trailing-edge process technologies. But that doesn't mean, he says, that the leading edge can be left up to others.
'It is fundamental that Europe increases its presence in manufacturing because it is fundamental to the business,' said Marco Rossi, group vice-president for strategic planning at STMicroelectronics. 'For More than Moore, the value is in the verticality, reaching through technology to services.
'If you want to develop anything built on top of process, you need a tight relationship with manufacturing and you need a critical mass of people. The concept of the pilot line is over. You either produce or you're pure fabless.'
Manufacturing in Europe might have waned in the past 20 years, but investment in production has returned. STMicroelectronics is extending the life of its Crolles fab. GlobalFoundries has decided to increase production at its Dresden facility, providing a potential new impetus to leading-edge production on the continent.
'GlobalFoundries is an important player in the ecosystem. Is GlobalFoundries in Europe key? The answer is definitely yes,' said Rossi.
Others are less sure. Bagherli argues that you can get equal access to bulk CMOS and differentiated processes from anywhere in the world, so there's little differentiation to be had from local production.
'We can build a success story using less than leading-edge processes by differentiating through design,' Bagherli argued.
Boland said: 'It seems to us that manufacturing expertise is shrinking quite rapidly and, at some point, it will become unsustainable.'
But local manufacturing has its advantages, Bagherli said: 'The best people we get are the ones who worked in manufacturing before they joined us.'
Joep van Buerden, CEO of CSR, said: 'I feel very comfortable with the people we have because they will train the next generation, so in 20 years' time we will still be OK.'
But one commenter from the floor of the debate had his doubts: 'Wafer fabs provide highly trained people to start-ups and the fabless community. If we kill off manufacturing here we cannot feed these people into the start-ups.'
Will Europe recognise the importance of its semiconductor industry and step in to save it? Kundert at SEMI doubts we'll see much more direct investment in manufacturing: 'Bad experiences such as Qimonda have left a lot of blood on the walls, which means it then takes a lot of convincing the public policy level to invest in manufacturing.'
Michel Hordies, scientific officer at the European Commission, presented what may be positive news for the industry. The Commission has a plan to sustain what it calls Key Enabling Technologies, such as nanotechnology, nano-electronics, photonics, advanced materials, and biotechnology, as part of its overarching vision for EU2020, which calls for smart, sustainable and inclusive growth.
'The semiconductor industry is innovative, extremely competitive and the basis of the knowledge society,' he said.
Europe has been investing in the semiconductor industry, in one way or another, for years, and some question the return the Commission has had on that investment. Yet its indirect actions, such as legislating for a pan-European standard for digital mobile cellular telephony, have created multi-trillion Euro industries that have remained firmly rooted here for decades.
One questioner from the floor asked whether Europe should use the carrot more than the stick more often, imposing new standards in a bid to create local opportunities.
'We don't use the word impose,' said Hordies, 'we call it lead markets. But there are many channels through which public procurement can create substantial initial markets.'
Wrapping up the debate, Baillie pointed out that actions such as requiring a move to LED street lighting in five years, or demanding the use of smart metering in homes, could create substantial opportunities.
'Overall, though, we need a full and balanced ecosystem with access to manufacturing expertise, directly or indirectly,' he said. 'And lead markets can create discontinuities that create opportunities.'
Rossi added his own coda: 'Europe is like a great restaurant with a great chef but not very many tables. It's a question of will.'