In the context of burner reactors (the only fuel cycle that has ever been demonstrated for a full fuel load and the only cycle with any serious proposal for a new reactor).
The amount of uranium the industry thinks they might be able to find (not the stuff already found) before the fuel alone costs more than renewable energy is about 10 million tonnes. Bear in mind the ore for the lower end if this holds so little uranium that you get less energy per kg of material processed than you do by digging up coal.
Each kg of natural uranium produces about 140GJ of electricity in the current fleet or 80-120GJ in an SMR (which is the main proposal for expanding generation).
Current world primary energy is about 550EJ/yr. Electrifying could reduce this to 300EJ, but demand is also increasing.
If you dug up all the known and inferred uranium reserves today and put it in SMRs like a nuscale or last energy one to produce 10TW (the average annual energy goal for renewables), it would run out halfway through 2025. It wouldn’t even be enough for a full initial fuel load.
If it were all EPRs and AP1000s (which have an amazing construction track record) and no demand growth was provided to offset efficiency gains if electrification, you might squeeze a decade out of it.