Good Intentions. But honestly I doubt the outcomes. Atleast its good news for NXP/INFN/ST/Atmel and their employees.
Doing "20 percent of global chip production." competitively, in Europe , is a highly ambitious target.
Indeed - that's where Giorgio Cesana, FD-SOI mover & shaker at ST got his degree. They have a big analog/power team there, too (see http://www.advancedsubstratenews.com/2011/04/smart-power-saves-power/ , for example).
I am sorry, but I fear Europe has missed the boat. Even billions of euros from the tax payers will not change that. Anyway, I believe in the future of the software industry, IT consulting and services. So, let's spend the money better for those projects.
You might be right concerning 'burning' tax money, but do you think that there is no growth in the chip industry? IT consulting as European strategy??! How many employment opportunities is going to generate?
a little too late for europe, a new leading edge 300mm fab is 6 billion dollars, 450nm I dont even know how much, so three of them? I dont think so, especially when only two foundries are making money, TSMC and SMIC everybody else is losing money.
I don't think there is a revenue model....
apart from tax and spend....
But to quote from the "Nanoelectronics beyond 2020" document: "The companies and institutes in Europe’s nanoelectronics ecosystem propose a strategic research and innovation program with a total investment
of 100 billion € up to the year 2020. By 2020, the programme aims to increase Europe’s
nanoelectronics-based world-wide revenues by over 200 billion € per year, and create an
additional 250,000 direct and induced jobs in Europe."
It all starts with convincing young people that it doesn't make sense anymore to pick an education aimed at earning "easy money" or get a diploma in a relaxing way. We have more than enough unemployed communication advisors, lawyers, economics etc.
A technical profession is still considered difficult, low status, geeky, boring etc. People forget an economy can't sustain in the long term by just shifting "services".
I do hope EU considers setting some of the money aside to create micro-seeds for chip startups. Big grants often leads to big waste. Putting aside 5% of the $6B program to fund one hundred early startups would give incredible long term ROI. $3M should be enough to go from concept to low volume production in many instances.
Bottom line €6B would not even pay for one wafer fab at the geometries relevant over the next 7 years and 700M/yr over 7 years is a drop in the ocean.
What is required if the EC wants to account for 20% of global wafer starts is something like the Mega project in the mid 80s along with structural changes in the tax regieme to make it viable to build wafer fabs in Europe
Also VCs are not putting money into fabless companies irrespective of whether the EC is one of their LPs or not
Andreas knows this very well
The only thing that will fix the semiconductor market is to feed large amounts of the tax take on semiconductor enabled products back into fabs and fabless companies
Don't hold your breath
It is because VCs are not putting money in to semiconductor startups that the European Commission is getting involved, but as indicated before that is being done under a different program due to start in 2014.
A good proposal. Atleast the EU is taking some steps. As adapteva has said, a sizable portion of this fund should go towards funding high quality startups. Even if they build a single fab, that should do I suppose.
I think Europe should focus on semi product design not manufacturing. The vast majority of semi companies are fabless for good reasons. I suspect a lot of the US semi production is a actually done in Asia anyway. A previous comment stated only SMIC and TSMC are able to make money from semi production. More EU money won't change that significantly.
There is an argument that once the manufacturing leaves the design follows.
We are definitely seeing the rise of Greater China as a significant design entity since chip manufacturing started to migrate there. Meanwhile we are seeing European design start to struggle. Note the linked examples of Nokia and ST-Ericsson.
I hope something good arise from the proposed investments. I'm into hardware and want to stay in europe.
It will get tricky though. The big players (ST, NXP, Infineon) have a strategy of outsourcing leeding edge manufacturing to Taiwan. Infineon fabricate power semis successfully, ST analog/MEMS successfully and NXP is reducing their debt. Neither of these really need the most modern facilities to do that. They can basically fabricate what they specialize in in legacy fabs while letting the asians do the heavy tech investments in capacity for digital. That's the whole idea behind the fab lite strategy.
The big players really would have to realign their business models if europe is supposed to increase their share of manufacturing capacity. I don't see that happening, unfortunately.
What happens with design is that the good ideas and good engineers tend to relocate to the US (better financing, it seems like), feeding the US high tech sector. Guess that could change but it's unlikely.
Anyway, there is no fundamental reason why we shouldn't be successful in high end electronics so I hope we will all be surprised when we look back in ten years.
Innovation, Creativity, Information and Education are the only way that anything like this will succeed. Already there are too many industry "geniuses" that have led us down the garden path with innovations that will go nowhere or have gone nowhere (like EUV (Hey 15nm is still X-Ray)) that are certainly sucking a lot of money away from anything meaningful like is being suggested. If Europe wants to pull this off, they will actually have to put their pride on the shelf and actually cooperate with each other in a meaningful way. Not at the executive level but at the technology level. Put the money into education and the creation of the next generation of engineers who are all ready to 'become'. They have grown up in the age of "information overload" and know how to handle it better than those of us who started in this industry way too many years ago. Companies should 'sponsor' or pay for the education of the best and the brightest and then put them to work on the future as soon as they graduate instead of letting them flounder in the abyss of unemployment because companies are to cheap to pay them, too cautious to hire them and so interested in today's bottom line that they lose sight of their futures. Through cooperation comes innovation and new ideas are born from education and creativity. As a US chip guy, I would like to see Europe pull it off, but I think the 500 pound (Chinese) Gorilla has already gotten a huge head start on what is being suggested and are not going to back off any time soon to let others take their share of the prize. It may be too little too late but that doesn't mean Europe should give up. They just need to be more creative in the way this is planned and the way it is executed; not business as usual and 'same old - same old" as has been the case for years.
European engineers usually don't. At least that's my experience.
Anyway, leading edge fabs are highly automated which means that factors like cost of capital, up time for the machinery and infrastructure, strict process control and taxes probably influence the profitability more than the wages of the process technicians and engineers.
The main reason that TSMC is TSMC is because they seem to manage those parts better than anyone else. That europe/Japan/US besides Intel fail to compete in those terms in leading edge digital is where the problems are located. Problems that unfortunately is much more tricky to fix than making the fab-workers into serfs.
Blog Doing Math in FPGAs Tom Burke 4 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...