Early stage technologies especially something as capital intensive and politically sensitive as nuclear are almost always messy in the beginning. NuScale Power was never going to be a straight line up, and anyone buying into SMRs should have understood that this is a long duration, high risk, high reward play.

The criticism around cost overruns and scalability isn’t wrong, but it’s also not the full picture. SMRs were never meant to replace large scale generation for entire cities. Where they do make sense is exactly where you pointed targeted, high demand, always on use cases like data centers, industrial sites, or remote grids. That demand is only going one direction with AI and cloud infrastructure expanding rapidly.