LONDON – Anti-IMEC sentiment across Europe has been putting the continent's role in the transition of chip production to 450-mm diameter wafers at risk, according to Mike Bryant, analyst with Future Horizons Ltd. (Sevenoaks, England).
Already Europe's role is somewhat limited with inward investors Intel and GlobalFoundries as the only companies likely to make chips on 450-mm wafers in Europe, But ASML Holding NV (Veldhoven, The Netherlands) has become almost a monopoly supplier of lithography equipment for leading-edge production nodes and stands to benefit from the transition.
Bryant said that a lack of vision and funding in Europe prior to 2011 led to the Global 450 Consortium (G450C) being based at the College of Nanoscale Science and Engineering (CNSE) in Albany, New York. "Europe has got to find a complementary role in 450-mm R&D now," said Bryant speaking at a one-day meeting here on Thursday (July 12).
Bryant said that his investigations showed that Intel and GlobalFoundries had been agnostic about the location of the wafer R&D and would have been happy to work at IMEC (Leuven, Belgium), a world-class research institute on microelectronics, rather than Albany. "A $200 million commitment would have killed it [G450C in Albany]," Bryant said. As it is IMEC and Europe's leading chip equipment makers are going to have to participate in Albany and may not find it easy to draw on European public funds to "create jobs in the U.S," Bryant added.
Bryant also painted a picture of the battle lines the consultancy had to cross as it researched a report on the future of 450-mm chip manufacturing that it prepared for the European Commission in 2011 and which was recently published (see European report considers 450-mm More-than-Moore fab).
On the pro- side of the 450-mm argument are chipmaking equipment manufacturers such as AMSL and ASM International and IMEC (Leuven, Belgium). On the other side are Europe's leading indigenous chip companies STMicroelectronics NV and Infineon Technologies AG and research institutes such as CEA-Leti in Grenoble and the Fraunhofer Institutes in Germany, Bryant said.
"European report considers 450-mm More-than-Moore fab"??? This makes no sense at all. More Moore requires 450-mm, perhaps. More-than-Moore clearly can be done on 300-mm wafers, if not on 200-mm, it is not the size of the wafer that matter here! Kris
One comment (to date) has been removed from this forum.
While we encourage freedom of speech, EE Times reminds contributors that they should not make personal, religious or racial slurs. Where they are brought to our attention they will be removed.
Regardless of who is "right", I have one pertinent question:
Why should the Flemish taxpayer subsidize 450mm R&D at Imec, as they are doing, when no European semiconductor company is interested in this?
It makes no sense. In Flanders, the government seems to think that innovation has been accomplished as soon as the subsidies are distributed (preferably to Imec). The world must think that we are either too rich or crazy.
If Intel, TSMC or Samsung wants it, let them pay. ASML is showing the wise way.
''...STMicroelectronics has been pressing for money from the European Union to help it upgrade 200-mm wafer fabs so that they can make chips on 300-mm diameter wafers, Bryant said...''
As an employee of STM, I find it surprising. Can anyone address me to the sources for such a statement ?
Second, looking at the original report, again issued by Mr. Bryant's firm, "European report considers 450-mm More-than-Moore fab", I used to consider 450mm associated to More-Moore, rather than More-than-Moore. Too simplistic? Could anybody show me already a fully 300mm More-than-Moore, loaded fab, to start with? In my (little) knowledge, TI is the only declared owner of a 300mm ''analog'' fab, supposed closer to my personal definition of a wannabe More-than-Moore fab.
Third, any article depicting any strategic choice as fruit of some 'sentiment' against someone else may be perceived as the result of the 'sentiment' of the writer vs. those that don't share his/her own vision. Bringing the debate to the realm of personal opinions is not what people would expect from an accounted consulting company.
Fourth: ASML business is founded on More-Moore, and 450nm fabs are a part of it. IMEC, whose dutch labs are at less than 15 min car drive from ASML HQ, has historically leveraged on its proximity with the litho giant. May anyone infer that interests for ASML and for IMEC may be somewhat entangled? The bigger customers for ASML are INTEL, SAMSUNG or TSMC, just to name a few, so european manufacturers may just wish to stay more focused on actual More-than-Moore (like MEMS) opportunities rather than venture in astronomical efforts for below 22nm technologies. ST or Infineon don't have enormous asset in 300mm, why should they invest now in 450? This is my (simplistic) Present Vision. For Future Visions, Mr. Bryant will take care.
"Regrettably STMicroelectronics, NXP and Infineon are now no longer tier-1 companies"
What a short-sighted statement. All 3 companies are tier-1 companies - in MEMS. Just because they may not be at the bleeding-edge of memory or logic manufacturing doesn't mean that they aren't leaders. The semiconductor industry has evolved - it's more than just about memory and logic these days.
Intel has reportedly said it will ONLY by 450-mm equipment from companies that are engaged with Global 450 Consortium and putting equipment into Albany for trial.
I would expect companies such as Tokyo Electron to be engaged.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.