Render vs other idle gpu grids

I’m trying to understand the difference between all these idle gpu projects. Let’s take Render and Nosana for example. And please correct me where I go wrong.

After researching both projects Render and Nosana, it appears they hold the same concept by creating a grid of idle gpu you can pull from. However, Render markets toward helping render images and videos, and Nosana markets their grid for helping with AI power.

If this is true, why wouldn’t who ever creates the largest grid rent out their gpu for any purpose? Why market toward a certain sector?Just confused on the potential for each project as I discover more of them like IO.net to name another.