Thursday, December 14th 2023
AMD Releases FSR 3 Source Code on GPUOpen
AMD on Thursday announced the first release of FidelityFX Super Resolution 3 (FSR 3) source code through the company's GPUOpen initiative. The company just set up an FSR 3 source code repo on GitHub that game devs everywhere can take advantage of. This includes the complete source for DirectX 12, and the source of an FSR 3 Unreal Engine 5 plugin. With it, the company also released extensive documentation that helps developers understand the inner workings of FSR 3, so they could better integrate the tech with their games and applications. With this announcement, AMD also unveiled FSR 3 support for even more new and upcoming games, which include "Black Myth: Wukong," the three latest titles from the "Warhammer" franchise, including "Darktide," "Space Marine II," and "Realms of Ruin;" "Starfield," "Pax Dei," and "Crimson Desert."
Source:
GPUOpen
50 Comments on AMD Releases FSR 3 Source Code on GPUOpen
also the drivers have become pretty decent the last few months, only took them 2 years, could have been worse, or could it?
The in-house close solution might receive better 'kick-off' and initial development.
But once its marketing potential is exhausted and no longer generates significant profits, it will soon be forgotten and left in the corner.
While the open solution still have chances from other interesting parties.
Nvidia tends to innovate new features, lock them to their hardware and get a generation or two of exclusivity, then AMD benefits from following that and developing a decent open source version.
I don't really mind the cycle tbh, I am a big fan of innovation and new features, I don't want to play games with the same rendering tech ad infinitum, and it gives good options for those who either want to be at the bleeding edge and pay for it, or wait if they choose.
This is the way of all adoption, and both Nvidia and AMD know it and happily take their place in the queue each and every time.
Give us the 'Freesync result' and all is well, honestly, with this whole upscale and FG affair. The moment it gets there, is the moment I'll believe in it. Until then, we just don't know - it could die any day, or live for ten years.Nobody said anything about rich and popular. The subject is 'best', and open source does have some qualities to it to apply just for being open source. Still though, closed is fine, but with great power comes great responsibility. MS and Apple are both companies that haven't tarnished that responsibility all too much. They make mistakes, but they also make an effort to stay on the good side of things, to good effect, too. Especially MS, imho; Apple is more of playing the arrogant card, effectively calling every bug a feature and part of being in the cool Apple club. They are both however definitely open to feedback and act upon it. By necessity, but also simply because needs and goals align. They want Windows in our homes. They'll cater to the needs. Its a shame their cloud services are so expensive, but prior or without that? I can't imagine a PC world without MS and its attitude on things, I mean, they've never been super anal about DRM or licensing cost for what is a superb OS. Windows was always a free for all and it still is.
And at the same time, Linux on the other end managed to evolve much the same, up to and including a presence on about 3/4th of the world's phones.
Nvidia has and participates in countless open-source projects the issue is that you are not developers or familiar with that aspect so you just see DLSS as closed-source, which would make 0 point being open-source since it relies on a hardware AND on a ML model (so you'd need the whole training set etc to have all the source, good luck).
developer.nvidia.com/open-source
The tensor RT sdk IS open source so are many nvidia sdks, yet, even after all those years some people fail to understand that DLSS relies a lot on HARDWARE, there is no real advantage to make it open-source or cross-platform when it's tied to your own architecture AND training/ML models !
AMD makes FSR open-source to drive adoption and throttle DLSS expansion, it is because it's advantageous for them. Many other AMD components
AMD "anti-lag" component, the one that generated VAC ban on CS2 is closed-source, no one makes a fuss about it...windows and ios are not open-source, neither are most AMD components including windows drivers however:
- Swift
- Typescript
- visual studio code
- dotnet core (C# more generally)
- Both MS and Apple contributes to many open-source projects
Those are some extremely popular open-source components by your aforementioned companies, what's the difference between releasing an API for upscaling made to counter DLSS as open-source and a whole damn language like C# used by thousands companies ?It's a driver level feature, how could it be open source lol.No, that's what Nvidia claims, there is no way to know that for sure because it's not... open source. They've made claims like that which proved not to be true before, like saying that frame generation cannot work on GPUs other than 4000 series because the optical flow whatever was too slow without dedicated hardware and that turned out to be a straight up lie because AMD got it to work on compute shaders which run on everything.
So suffice to say I for one do not believe that.
I would focus more on the fact that company that by many is considered to make bad software for their hardware managed to do FG without specialized hardware. A feat that nvidia claimed is impossible and it needs Optical Flow Acceleration, that's why it's not available for older cards.
In case of FSR2 it is looking worse than DLSS (hope that after they have finished FSR3 they can focus more on upscaling part) but for Frame Gen now after they fixed VRR I do not see any downside compared to DLSS Frame Gen. And I used both.
Oh wait, I get refresh rate frames at 4k, never mind. :)