Tilemap performance

I am new to Phaser and experimenting with tilemaps, and I was wondering about the CPU usage to render them. For example, using Tiled, I made a 256x256 (32 px square tile) tilemap. I then loaded it in to Phaser (3.55.2) and when I zoom all the way out to view the whole tilemap my CPU is sitting at 10-11% usage.

If I load an older Phaser version (3.17.0) and create a static tilemap, zoom all the way out again, my CPU is around 1-2% to render it. I know they consolidated the static and dynamic tilemap stuff, but does that mean you don’t get the performance benefits of a static tilemap if you want it?

As another option I exported my tilemap from Tiled as a 8192x8192 PNG and loaded that in. Zooming all the way out to view everything has my CPU at 0%.

If I don’t need the flexibility of a dynamic tilemap and want better performance should I just stick to exporting my tilemap as an image? Or am I missing a configuration option somewhere?



I think there is no “static” tilemap rendering like in ≤ v3.24.

The 8k × 8k image might be too big for some devices. But maybe it’s fine for desktops.

Thanks for the reply. Glad to know that the static tilemap is gone for sure.

Yeah, the 8k x 8k image is bigger than I would go with it, just using it for testing.


hi, I’m interest in what you did in the end, do you advise using tilemap or just image placement?

I don’t have anything finished yet, but I will probably try it with just an image. If that becomes too big a pain I may just end up dropping back to an earlier version of Phaser.