Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Incorrect data in RGP's Workload and Wavefront distribution sections for NGG shaders #66

Open
Venemo opened this issue Apr 25, 2022 · 0 comments
Labels

Comments

@Venemo
Copy link

Venemo commented Apr 25, 2022

When comparing the number of vertices specified in a draw call vs. "Shaded vertices" and "PrimS wavefronts", the numbers don't make sense. Here is an example screenshot with the problematic numbers highlighted:

Screenshot from 2022-04-25 12-40-04

  • The highlighted draw call has 11904 vertices
  • Wavefront distribution claims there were 127 primitive shader wavefronts
  • The shader was running in Wave32 mode according to Pipeline state (not included in the screenshot)
  • But, according to the Workload section, there are only 127 shaded vertices

These numbers are clearly incorrect:

  • The number of shaded vertices should be equal to the first parameter of the draw call (minus vertex reuse)
  • 127 wavefronts in Wave32 mode are only capable of processing 127*32=4064 vertices, so the number of wavefronts also can't be right

Here is another example:

Screenshot from 2022-04-25 12-53-48

On this screenshot, we see a draw call with 6 input vertcies (2 triangles), but RGP claims that the number of shaded vertices is only 1, but it created 8294400 pixels from that 1 vertex? Which also makes no sense.

@chesik-amd chesik-amd added the bug label Jul 12, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Projects
None yet
Development

No branches or pull requests

2 participants