there will be lasting change to to our lives and lives of people in our oribt. I don't think we appreciate that vast majority of the workforce (in transport, hospitality, retail, manufacturing) is not changing at all.
And the continued move within the enterprise towards using cloud computing and SaaS products.
No matter which way you look there is an increasing demand for silicon.
I believe that the bigest drivers for silicon demand will be IoT devices, including automobiles, especialy when they become more affordable in the 3rd world.
The pressure from SaaS and cloud is more predictable than the consume market, sonthe semi industry can plan ahead, IoT not so much. Just see how the auto industry was caught by surprise by the increased demand.
The way we work and learn has changed, and it doesn't look like a reversion to pre-telework habits will ever fully occur. Several companies (including Salesforce, Twitter, Square, and Spotify) have said that telework will be a significant part of their post-pandemic reality.
Meanwhile, electric vehicles are slowly but surely taking market share from internal combustion engines that use less than a third as much silicon.
It feels like demand creation to me, and that means we ought to treat the prior capex roadmap for the industry as insufficient.
Even in the absence of demand creation, fab concentration in Taiwan is increasingly a liability in a world where Chinese economic hegemony is growing. One way to mitigate future perturbations that may result from military conflict, natural disasters, or other interruptions in Taiwan's fabs would be to build greater capacity outside of Taiwan.