Hello everyone !! Small (maybe ignorant) question, I believe prisma doesn’t use SQL JOINs under the hood, but separate SELECTs, can this represent a performance issue at big scale?

Doesn’t Prisma allow you to toggle that behaviour? In my experience, JOINs were faster for smaller queries but as the query becomes more complex and more tables are added to the JOIN, it becomes a CPU/memory bottleneck on the database instance. Switching the bigger queries to use multiple separate SELECTs were much more scaleable because the smaller SELECTs are more cacheable and it pushes the computation logic to the application instead of the database so less CPU/memory usage.