| 19 Jul 2025 |
 olivia | hmm | 07:19:25 |
| olivia | as an application dev I would be terrified of weird artifacts there | 07:19:33 |
| olivia | with future versions that I haven't tested with | 07:19:40 |
 K900 | There's also the whole "DXIL has no vendor extensions" thing where they have to basically smuggle FP8 coopmat into the shader compiler | 07:20:05 |
| K900 | But that's just DXIL being DXIL | 07:20:10 |
| olivia | I guess this bit is also surprising because I'd expect that you don't want to do the upscaling stuff on UI? | 07:20:21 |
| K900 | And they'll probably switch to native instructions for this once those actually exist (...right?) | 07:20:34 |
| K900 | Yes | 07:20:39 |
| olivia | so I had assumed the normal pipeline looks like "render all your non-UI stuff at low res" -> "feed to the upscaler" -> composite with all the UI | 07:20:45 |
| K900 | That's how it usually goes yes | 07:20:52 |
| olivia | and the driver can't tell the difference | 07:20:54 |
| K900 | But with driver injected FSR2 it just runs on the entire frame | 07:21:03 |
| K900 | Which is not great | 07:21:09 |
| olivia | yeah that's fucked | 07:21:14 |
| K900 | But it kinda works and they get to advertise a feature for MORE FRAMES | 07:21:18 |
| olivia | I just learned that fact a couple days ago when reading about fsr4 is my mind is boggled | 07:21:33 |
| K900 | I still remember magic D3D9 fourccs | 07:21:47 |
| K900 | It's just always been like that | 07:22:00 |
| olivia | yuck | 07:22:18 |
