PhysX and Flow GPU Source Code Now Available

[Kaotik]

Maybe with a compatibility layer, like Proton. A contributor to dxvk-nvapi, an API that provides Proton with the ability to use AMD technologies like DLSS, Physx, Reflex, etc.

Mirror's Edge PhysX does not use the GPU. · Issue #277 · jp7677/dxvk-nvapi

I installed PhysX via Protontricks, deleted the bundled dlls according to PCGamingWiki, and even threw in the Nvidia-Libs project. (It should be noted, that none of this changed the performance wha...

github.com

Actually there kinda already is part of that. FakeMichaus FakeNVAPI is included in OptiScaler allows to turn Reflex into LatencyFlex or AMD Anti-Lag

GitHub - cdozdil/OptiScaler: OptiScaler bridges upscaling/frame gen across GPUs. Supports DLSS2+/XeSS/FSR2+ inputs, replaces native upscalers, enables FSR3 FG on non-FG titles. Supports Nukem mod for DLSSG-to-FSR3 FG.

OptiScaler bridges upscaling/frame gen across GPUs. Supports DLSS2+/XeSS/FSR2+ inputs, replaces native upscalers, enables FSR3 FG on non-FG titles. Supports Nukem mod for DLSSG-to-FSR3 FG. - cdozdi...

github.com

 

[Sega_Model_4]

Now I noticed that my comment was half-finished. What was missing was to say that a contributor to dxkv-nvapi said that perhaps Proton could, in theory, get around this Physx issue.

 

[trinibwoy]

I wouldn't really compare publically available modern gfx APIs (which are very much still designed around multiple hardware vendors) to a singular vendor compute-centric API like CUDA (doesn't really expose fixed function HW too) in terms of potential surface area with respect to buggy implementations ...

We’re discussing a potential DirectCompute implementation which is no more vendor centric than DirectX. My comparison was to the significant per-application tuning that’s done for DirectX applications. DirectX is also a higher level API than DirectCompute and therefore has a higher surface area for optimization and/or bugs.

When gameplay code and physics code are heavily interwined between each other, any driver bugs can become a much more invasive showstopper to shipping your game on more device configurations.

Sure but it doesn’t need to start there. There are tons of ornamental effects that can run decoupled from gameplay.

 

[Lurkmass]

We’re discussing a potential DirectCompute implementation which is no more vendor centric than DirectX. My comparison was to the significant per-application tuning that’s done for DirectX applications. DirectX is also a higher level API than DirectCompute and therefore has a higher surface area for optimization and/or bugs.



Sure but it doesn’t need to start there. There are tons of ornamental effects that can run decoupled from gameplay.

A 'supplemental' GPU-based physics library could be viably considered if we're just sticking to Direct3D/consoles but how much customer interest would there be for such a self-contained solution that won't affect the game design ? PhysX had tons of non-trivial integration issues that stopped it's GPU implementation from being widely supported even just for purely visual/simulation effects because there still HAD to be SOME interaction/invasion (particle/body collision & vector force fields) with the in-game world to get convincing results ...

 

[trinibwoy]

how much customer interest would there be for such a self-contained solution that won't affect the game design ?

I’m guessing a lot more interest than for barely noticeable graphics upgrades that also don’t affect game design.

 

[Lurkmass]

I’m guessing a lot more interest than for barely noticeable graphics upgrades that also don’t affect game design.

Sure but it's easier to integrate these graphical upgrades (if you're on the same renderer) since they're more disconnected from the game design ...

 

[raytracingfan]

HWRT might increase interest in GPU physics. There's a lot of ray casting against bounding boxes, spheres, and capsules in physics engines, with HWRT you can trace rays against the actual full-detail meshes instead of proxy geometry. Doom: TDA is using it for hit detection.