7

Visibility rendering using traditional vertex/fragment shaders does 39 million tris in about 3.6 ms

With my newest renderer I can push 314 million triangles in about 6 ms right now

And this is just visibility, factoring in material evaluation of traditional deferred it would be at least like 10x worse. Meanwhile everything expensive about materials is completely independent of geometric complexity in my renderer

Literally me rn: https://youtube.com/watch/...

(cant include image because devrant doesn't want to)

Comments
  • 1
    Sounds amazing. Is it for work or private use?
  • 2
    @retoor For am upcoming game of mine. Because trying to make a AAA quality engine from scratch is totally a smart move...
  • 1
    @12bitfloat haha sure, if you're good enough. I'm making a regex parser from scratch - not sure yet if it was a good move :P I want to beat C builtin regex. Quite challenge.

    You're also writing C right?
  • 1
    I'm mostly writing in Rust. It's pretty nice but comes with it's own set of challenges

    Writing a regex engine is pretty cool. I'm not too knowledgable about it but some implementation details I heard sounded pretty insane :D

    Hows your progress so far?
  • 1
    Few questions:

    1. How does that measure up against engines?

    2. If cutting engine or an improvement over brand-name engines, what did you do differently?

    3. What kind of game are you making?
  • 0
    @Wisecrack Haven't tested it precisely but it should be a magnitude or so faster than traditional engines and a bit faster than UE Nanite (which doesn't seem to be all that efficient). Keep in mind this is just about getting pixels on screen. Shading and lighting those pixels will take about about the same time as traditional

    The reason I can achieve this is mostly because I know quite a bit about low level graphics development and because it's made from ground up based on modern tech. I don't have to shoehorn visibility rendering into a 20 year old engine that was never made for that

    But it's also about momentum. A lot of studios don't want to (or can) completely change everything, which is why they are now instead switching to mesh shading. It's different but not /that/ different. It's also just bad. I played around with it and came to the conclusion that mesh shading is a waste of time, which led me to compute rasterization
  • 0
    @Wisecrack I'm making a photorealistic open world racing game :D

    That also helps a lot knowing what I want the engine for, since I can now tailor it for my exact needs
Add Comment