How much can we scale up production in the compute supply chain?

How quickly we will be able to increase the production of hardware and the required computing infrastructure (such as data centers) as we increase the investment in compute for AI?.
To what extent could ASML, TSMC, and other suppliers across the compute supply chainscale up their production on a 1/5/10 year timescale if the prices of their products went up dramatically (eg x10)? What are the key bottlenecks for such a massive scale-up across the compute supply chain?