450 mm wafer size is an off and on discussion topic. More thinking and planning should be happening just about now. What is the current sentiment?

This is really a business question. There are a couple of basic considerations.

First, there is cost. The cost per wafer can be divided into energy/material cost (which is proportional to area), wafer handling cost (which is proportional to wafer weight or volume), and footprint cost (real estate, roughly proportional to wafer area).

Then there is number of products per wafer. It is roughly proportional to area.

As a result, the cost per product is almost the same, and maybe higher due to the handling of thicker wafers. This is not including the initial development and installation costs.

So the cost is not the real reason to drive wafer size.

You also have capacity. The ability to turn out so many wafers per month is key to the company's business. A company like Intel may have four 300 mm fabs running actively at a given time. If it had been using 200 mm fabs, it would have needed to keep maybe 8 fabs running actively at a given time. This is pretty complex and would have required a lot of staff, ramp coordinations, etc. On the other hand, maybe only two 450 mm fabs would satisfy the capacity. The two fabs would not come up all at once; one fab would be running for a good part of a year. Which brings up the greatest risk: what if that fab had an excursion or went down?

So the current usage of 300 mm fabs is about the right balance of low risk and low complexity.