Much better, looks just like Native, and doesn't require game patching, so no more waiting, as Nvidia will just have it in their Game Ready Drivers, so hopefully take off now, and about time Should have been like this at the start though tbh
This is how it should have been from the start. Requiring devs to update their game was a terrible idea imo. The name is still stupid though. It really is upscaling with sharpening. Nice to have but something i would try to avoid personally.
Well one part of it is actually downsampling, but no one "dared" to use that. DLSS 2x " DLSS = Performance Mode, and DLSS 2X = High Quality Mode DLSS 2x that seems to render at Native resolution and then does more complex DL network operations for much higher quality: "DLSS 2X. In this case, DLSS input is rendered at the final target resolution and then combined by a larger DLSS network to produce an output image that approaches the level of the 64x super sample rendering – a result that would be impossible to achieve in real time by any traditional means. Figure 21 shows DLSS 2X mode in operation, providing image quality very close to the reference 64x super-sampled image." The DNN AI also has to intelligently combine inputs based on its 'understanding' of the motion of the scene, instead of strictly following motion vector values like TAA, which gives us results like this: " towards the end of the article https://devblogs.nvidia.com/nvidia-turing-architecture-in-depth/
Yep, I'm annoyed it isn't in more games at the moment but I guess that will come with time. They need to DLSS all the things DLSS 2.0 is a game changer. 1.9 was like pushing a beta as an end product and made the tech look poor compared to it's potential.
From: https://www.resetera.com/threads/nvidia-dlss-2-0-a-big-leap-in-ai-rendering.176728/ https://www.nvidia.com/en-us/geforce/news/nvidia-dlss-2-0-a-big-leap-in-ai-rendering/ Faster, more customization options what sounds like a big one with network changes and possibly leading to more titles supporting DLSS and also improving the quality of it. EDIT: Coming to "Control" in a March 26th update too, should see some more about it then when there's a title out supporting it already and that should make it easy to compare and test also.
it seems like DLSS 2.0 has quite little to do with the native resolution DLSS 2X we were presented a couple years ago, looks like a complete rework of this solution ? Game agnostic is a big deal moving forward. I'm hopeful at least, DLSS has been rather underwhelming so far (hope BF5 upgrade DLSS to 2.0...)
Tbh didn't expect DLSS improvements like this before Amphere (which I guess will have further hardware optimisations as well). The "no game optimisation required" hopefully means a lot more games will support it as it's clearly now going to be much easier to do. I bet under the scenes they just have a bunch of standard profiles and the devs just pick the most suitable one like with SLi.
"Originally from https://forums.guru3d.com/threads/dlss-2-0-the-prodigal-son.431302/ " Looking forward to Control update featuring this updated DLSS, if any game should showcase an improvement this is clearly this one. Arguably their former approach of having to train their model 'in screen space' requiring training on full game picture and specific resolutions sounds a lot like a 'brute force' approach. I'm glad they moved away from it as it wasn't making any sense - DLSS should have been trained only on edges not full pictures: the work intended by the algorithm is to smooth out edges - being temporal or static - if you only train the model to solved edges, you don't need to be game specific nor resolution specific which appears to now be the case with DLSS 2.0. We've been stuck with TAA for far too long - this technology has a massive performance cost, creates ghosting and blurriness - it needs to go. Pretty confident AMD will come up with an alternative to DLSS. I wouldn't be surprised if DLSS end up being a Nvidia Control panel option in the coming months. Nvidia has been actively working with Unreal Engine for years now, with specific version of unreal engien available on github - so the number of game benefiting any of its technology has been growing steadily for a while now.
The other thing that's interesting about this is how much software dev work must have gone into this - it's basically taken Nvidia 3-4 years to get the software to a state where it's properly ahead of anything you can do without specialised hardware. Hence the software that makes this work is hard. Even when AMD/Intel do add the AI hardware, they are going to a long way behind in software.
I have to say this: who are those people being interviewed by in those videos? I mean, they are talking over either my left or right shoulder...so, who were they being interviewed by? If the answer is "no one", when...why were they not talking directly into the camera? If the answer is "style", then...why are they pretending to be interviewed? and perhaps moreso: why are they deliberately deceiving the person or persons watching the video? If the answer is "style", then, don't you think they look stupid?
sharing is caring? this defiently look more promissing then it first incarnation, but not for the price of admition my 1070ti is fine for what play
