Is there a way to make copy operations such as CopyResurce
(d3d11) and glBlitFramebuffer
(ogl) on multi-sample resources faster with Intel UHD integrated graphics?
For Direct3D11 I was able to collect some data (using Intel's graphics frame analyzer), and it looks like a single call to CopyResource
can take up to 6ms (both textures being ~1200x 800, r8g8b8a8_unorm, 8x).
- Is it a known issue?
- Does it mean that it's basically impossible to copy ms textures during the render loop or is there something one could do to make the copy faster?
- Could that be a driver issue? (I used version 31.0.101.2115 while profiling)