Which of these approaches is better for performance? (sprite related)

You have a top-down view game.
The floor/background is made up of a, let's say 15 x 15, grid of tiles, which all have the same "tile" image.

so you create 225 sprites.

The sprites don't move (or do anything except display their image), but the camera follows the player, because the 15 x 15 grid is bigger than the screen.

It occurred to me, that instead of having 225 sprites, I could draw the "tile" images to one large image big enough to contain the 15 x 15 grid. And just have one sprite.

But then I thought, this would mean I have a very large image in memory.

Is 225 sprites enough to cause performance issues?

A side question would be, if a sprite is not within the space that will be shown on the screen, is it still drawn? Does that even matter?

Yes, 225 sprites would be bad for performance.

One is much better, don't worry there's lots of RAM.

as an example of memory usage my sprite imagetable contains 1,980 frames, and uses about 4MB RAM out of the available 16MB

RAM. Memory is the cheapest, most plentiful thing on Playdate. CPU and refresh rate are more scarce.

You could also try the tilemap function if you need programmatic manipulation of the tiled background.

Off-screen sprites are not drawn but they still have other stuff going on that may affect performance.

Ok, thank you for the answer. Makes perfect sense.

Questions like this are great first tasks once you get a device. Answering them makes you feel closer to the hardware.

What you said about RAM being much more abundant than CPU power certainly seems like a key piece of knowledge.

So, thanks again.