Simple example of creating a balloon on Android

Hi All,
I want to replace a java class that draws a icon using opengl es with one using Vulcan. Are there any examples close to that somewhere?


It’s called “Vulkan”. And using it to draw a simple icon sounds like an overkill. There are no simple examples, because Vulkan is not a simple icon-drawing API. Anyway, you give insufficient information to offer any reasonable suggestion.

Thanks for the reply. I thought it was a replacement for opengl but as you said it was meant for something else. Probably why not much activity around it.

Good Luck!

OpenGL isn’t a simple icon-drawing API either.

Oh you are right about that. I was trying to simplify in words what I needed to do. But for rendering say 50k-100k simple icons I wanted to raster in parallel and move independently I needed something faster than the single threaded Opengl ES. I only needed that part of an existing opengl es program to render faster.
I know now that even with Vulkan I might only see a 2-3x improvement so I am looking elsewhere.
Again thanks for your replies!

I found some great articles by NVIDIA which are a big help in optimizing the code.
Again many thanks!

It is a replacement for OpenGL. That is, for its primary mission of tightly abstracting the varied GPUs. So some original OpenGL users are surprised, adjusting badly to Vulkan’s more explicit and low-level nature. Though it’s nothing new; same surprises happened in the quantum leap from OpenGL 1.0 to OpenGL 3.

OK, that’s more like it. Still, submitting stuff to a GPU incurs an inherent overhead. Do you have some measurement if OGL even helps vs just simple CPU implementation?

There’s couple of examples showing offscreen rendering, or mesh\scene rendering. But of course nothing that will particularly fit your niche needs. Not that I think copy-pasta random example will help you learn or use Vulkan effectively.


I say “only” because I saw an article for mobile that limited the performance gains based on the number of gpu cores. It was 1 thread per GPU core I believe. 2x improvement can be significant but only if the developer has exhausted all other opengl es optimizations first.

This topic was automatically closed 183 days after the last reply. New replies are no longer allowed.