Optimizing variable tile map rendering

[Shuvi22]

Hi everyone who stopped by here. My problem is that I want to draw a tile map of about 512x512 tiles. This loads the CPU by half and the FPS is about 5-10. If I draw every other tile, the FPS increases several times. Because each tile can be changed, I can't merge all the sprites into one, or merge everything into one sprite every time. I had the idea of merging sprites into chunks(i.e. the same thing as before, but only a small part of the sprite would be recreated into one), but I don't know if that would work correctly. How good is this idea and are there any better ideas?

[eXpl0it3r]

Are those tiles all visible at the same time?

Common strategies are:

• Using a vertex array and a single texture to reduce draw calls
• Using a quadtree (or similar structure) to determine what is actually visible on the window and only draw that
• If you're zoomed out a lot to view all those tiles, consider using lower-res textures, which then fit more easily on one texture to be rendered with a vertex array

Generally using a vertex array should yield good enough performance even with more complex tilemaps.

[Shuvi22]

Yes. Thank you. I recently learned about vertex array and this is what I need