And last year it was said:
“By the end of next year, we will have a better idea of how difficult it is going to be to move to 450mm,” said Paolo Gargini, a director of technology strategy at Intel. “By the end of 2009, companies will be able to map out the costs and how long it would take to recover investment.”
So, it appears a decision has been made to move forward without really understanding how difficult it will be. What if the costs and recovery period are too long? How is this cost mapping effort going and who is actually doing it?
The reality is that there is no such thing as a “simple scale-up” – without full diagnostics and data aquisition, intelligent equipment with on-board condition-based maintenance, higher-speed processing schemes, buffering schemes, and better factory layout and design (to name a few), 450mm will be a flop, and a very expensive one at that. And that’s before you deal with the physics. Architecting a 450mm fab is serious business that will take signficant time, energy, investment, and creativity. That might take a while, especially when not everyone has really bought in.