Nvidia has introduced an open-source framework to make it easier for developers to plug anyone’s upscaling algorithm into the rendering pipelines of their games. Intel’s on board, and AMD’s FSR 2.0 looks like a prime candidate, too. None of which seems like something I would have expected to be writing today. Or ever really.
Historically, Nvidia has not exactly been known as the open-source champion. Team green has generally been very much into the proprietary schtick for its developer-facing features, often requiring devs to sign up to use particular graphical goodies. Whether by design or no, that would potentially lead to the exclusion of other manufacturers’ competing tech, often purely by virtue of the fact that dev time is limited, and expensive, and sometimes you’ve just got to pick a side and stick to it to get a game out in time.
With Nvidia Streamline, however, developers no longer have to figure out a way to manually plumb every different upscaling SDK into their game. This has the combined benefits of shortening the amount of time it takes to add different solutions and providing more opportunities for both manufacturers and gamers to access those features.
Streamline has been designed as more or less a plug-and-play framework that exists in between the game itself and the final render API, be that DirectX or Vulkan. Yes, we’re even talking about it working with Vulkan, though that is in beta at the moment. Interestingly Streamline isn’t restricted to DirectX 12, either, as Nvidia knows the benefits of super sampling beyond ray tracing and is making the framework functional for DirectX 11 games, too.
The open-source nature of Streamline, now released in its entirety on Github, means that hardware manufacturers can create their own plug-ins for the framework. Intel has already made its own plugin for XeSS, so games will be ready once the Arc Alchemist GPUs finally are, and that means AMD can get in on the action as well.
At the moment, however, AMD’s FidelityFX Super Resolution (FSR) comes into play after the standard render pipeline, and so probably isn’t a good fit to run with Streamline. But FSR 2.0 is a different beast—now a temporal-based version, like DLSS and XeSS—and will actually come into the graphics pipeline. Which should also make it even easier to implement for developers.
Already AMD reckons devs will be able to implement FSR 2.0 in less than three days if they’ve already got their game up and running with DLSS 2.0. But if it’s simply a matter of adding another plug-in to Streamline that would be even quicker.
By providing a standard, open-source way for companies to easily get their super sampling shinies in front of more gamers, Nvidia is actually opening up choice for players, and not purely looking at locking them into the GeForce way of the world. Which seems counter to the common narrative levelled at the green team.
Anything that enables us to get more super sampling techniques into our games gets a big tick from me. I don’t want anyone to be in a situation where they’ve got one company’s card and are excluded from a free performance boost simply because they chose the ‘wrong’ manufacturer for that game. And this looks like another step towards eradicating that possibility.
So long as AMD signs up, I guess. It’s open-source though, so it’s going to need a very good reason not to. We’ve asked AMD whether it’s considering a plug-in and will update when we hear back.