I'd be interested in learning more about the European semiconductor industry's vision of how a 450-mm volume production fab would operate. Would there need to be a common platform/set of processes established across the major manufacturers in order to share in the benefits of the fab, and if so, are the leading-edge high-volume products produced by the lead European chipmakers like STMicro, Infineon, etc., conducive to that type of model?
Thanks for sending the link Peter. I just read your blog post and I think you're spot on. You bring up a good point that the economics for a larger wafer size for More-than-Moore applications don't make a whole lot of sense. I'll take a further look at the EU report too.
I just wondering which company in Europe will be able to establish a 450mm wafer fab. Looks like there have not been any new fabs built in Europe for the last few years. I can see only Intel and GlobalFoundries that are non European companies building new fabs in Europe (on top of those in Ireland and Dresden). ST, Infineon and NXP are just not capable of that financial burden, and with Europe financial crisis it is even more remote possibility.
Well the idea is that a consortium would build EuroFab450 and away from the leading-edge to keep the cost down.
Much of the cost at 450-mm is likely to be in the very advanced lithography required for leading-edge manufacture.
The shell of the building is nothing. The 450-mm handling will be relatively low volume and high cost to begin with but if you play away from the leading edge offering More-than-Moore process technologies you might be able to get the cost down considerably
How low? $500 million and $1 billion?
I just don't know.
Now the question is whether a consortium of European chip companies, including the likes of X-Fab and Lfoundry and the bigger players, can be persuaded to chip in to a 450-mm fab AND whether they can somehow get a commercial advantage from that enormous wafer size.
It has never been done before. It doesn't look likely. Europe could opt to do nothing.
And then we can look forward to flipping burgers for Chinese tourists visiting places of historical interest in Europe.
We will be flipping burgers if we no longer make (semiconductor or other) products that people want to buy.
Thinking that we will have the products just because we have the fab is absurd. Conversely, if we have the products it doesn't matter where the fab is.
What a EuroFab450 project would do is bring us closer to the flipping-burger point, by drawing attention and funds away from what really matters: building globally competitive private companies.
Blog Doing Math in FPGAs Tom Burke 16 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 ...