A mobile application that I’m working on is expanding in scope. The client would like to have actual 3D objects in a product viewer within the app that a potential customer/dealer could zoom and rotate. I’m concerned about bringing a model into an OpenGL environment within a mobile device.
My biggest concern is complexity. I’ve looked at some of the engineering models for the products and some of them contain more than 360K faces! Does anyone know of any guidelines which would discuss how complex of an object OpenGL is able to handle?
OpenGL is just a specification and doesn’t deal with geometrical complexity. BTW: OpenGL doesn’t treat geometry as coherent objects. For OpenGL it’s just a bunch of loose points, lines or triangles that it throws (i.e. renders) to a framebuffer, one at a time.
Any considerations regarding performance make only sense with respect to an actual implementation. For example a low end GPU may be able to process as little as 500k vertices per second, while high end GPUs process several 10 million vertices per second with ease.