Well there is a slight catch as of right now based on Xbox Series X RDNA2 specifications.

Tensor cores are very specialized cores that can accelerate INT-8 TOPS which is what's used to Render DLSS according to Digital Foundry. The thing is, the Series X and Ps5 and probably the Series S will have cores that can also do them however not as fast as Nvidia's Tensor cores. Now due to architectural differences and software differences and etc, maybe RDNA 2 might not need them to be as fast so who knows. I am not that technical into how DLSS works at a low level loll.

The point is though, with what we know now, if we were to port DLSS over to the Series X. It would take twice as long to Render DLSS compared to a 2060.

And there's a bit of an interesting article here too about it:

https://www.eurogamer.net/articles/digitalfoundry-2020-image-reconstruction-death-stranding-face-off

"There's an important point of differentiation between Nvidia's hardware and AMD's, however. The green team is deeply invested in AI acceleration across its entire business and it's investing significantly in die-space on the processor for dedicated AI tasks. AMD has not shared its plans for machine learning support with RDNA 2, and there is some confusion about its implementation in the next-gen consoles. Microsoft has confirmed support for accelerated INT4/INT8 processing for Xbox Series X (for the record, DLSS uses INT8) but Sony has not confirmed ML support for PlayStation 5 nor a clutch of other RDNA 2 features that are present for the next generation Xbox and in PC via DirectX 12 Ultimate support on upcoming AMD products.

Broadly speaking then, the Xbox Series X GPU has around 50 per cent of the RTX 2060's machine learning processing power. A notional DLSS port would see AI upscaling take 5ms to complete, rather than a 2060's circa 2.5ms. That's heavy, but still nowhere near as expensive as generating a full 4K image - and that's assuming that Microsoft isn't working on its own machine learning upscaling solution better suited to console development (spoilers: it is - or at least it was a few years back). In the meantime though, DLSS is the most exciting tech of its type - we're sure to see the technology evolve and for Nvidia to leverage a key hardware/software advantage. The only barrier I can see is its status as a proprietary technology requiring bespoke integration. DLSS only works as long as developers add it to their games, after all.

As exciting as the prospects for machine learning upscaling are, I also expect to see continued development of existing non-ML reconstruction techniques for the next-gen machines - Insomniac's temporal injection technique (as seen in Ratchet and Clank and Marvel's Spider-Man) is tremendous and I'm fascinated to see how this could evolve given access to the PS5's additional horsepower."

With that being said though, there are a lot more months for both Sony, MS and AMD to reveal more and more about their GPUs and other features so things could change.

Well there is a slight catch as of right now based on Xbox Series X RDNA2 specifications.

Tensor cores are very specialized cores that can accelerate INT-8 TOPS which is what's used to Render DLSS according to Digital Foundry. The thing is, the Series X and Ps5 and probably the Series S will have cores that can also do them however not as fast as Nvidia's Tensor cores. Now due to architectural differences and software differences and etc, maybe RDNA 2 might not need them to be as fast so who knows. I am not that technical into how DLSS works at a low level loll.

The point is though, with what we know now, if we were to port DLSS over to the Series X. It would take twice as long to Render DLSS compared to a 2060.

And there's a bit of an interesting article here too about it:

https://www.eurogamer.net/articles/digitalfoundry-2020-image-reconstruction-death-stranding-face-off

"There's an important point of differentiation between Nvidia's hardware and AMD's, however. The green team is deeply invested in AI acceleration across its entire business and it's investing significantly in die-space on the processor for dedicated AI tasks. AMD has not shared its plans for machine learning support with RDNA 2, and there is some confusion about its implementation in the next-gen consoles. Microsoft has confirmed support for accelerated INT4/INT8 processing for Xbox Series X (for the record, DLSS uses INT8) but Sony has not confirmed ML support for PlayStation 5 nor a clutch of other RDNA 2 features that are present for the next generation Xbox and in PC via DirectX 12 Ultimate support on upcoming AMD products.

Broadly speaking then, the Xbox Series X GPU has around 50 per cent of the RTX 2060's machine learning processing power. A notional DLSS port would see AI upscaling take 5ms to complete, rather than a 2060's circa 2.5ms. That's heavy, but still nowhere near as expensive as generating a full 4K image - and that's assuming that Microsoft isn't working on its own machine learning upscaling solution better suited to console development (spoilers: it is - or at least it was a few years back). In the meantime though, DLSS is the most exciting tech of its type - we're sure to see the technology evolve and for Nvidia to leverage a key hardware/software advantage. The only barrier I can see is its status as a proprietary technology requiring bespoke integration. DLSS only works as long as developers add it to their games, after all.

As exciting as the prospects for machine learning upscaling are, I also expect to see continued development of existing non-ML reconstruction techniques for the next-gen machines - Insomniac's temporal injection technique (as seen in Ratchet and Clank and Marvel's Spider-Man) is tremendous and I'm fascinated to see how this could evolve given access to the PS5's additional horsepower."

With that being said though, there are a lot more months for both Sony, MS and AMD to reveal more and more about their GPUs and other features so things could change.

Don`t forget the Unreal Engine 5 demo on PS5 at 1440p that Digital Foundry said they couldn`t really distinguish the pixels and that was basically making pixel count rather obsolete.

Also both MS and Sony have talked about half precision integer capability on their console that would support the temporary reconstruction technique.

Not to forget as well that PS4Pro made big use of single precision for their checkerbordered technique that since that time had DF saying their checkerbording from 1440p to 4k had almost the same level of clarity as native 4k for a fraction of the cost.

It is silly to think that everyone have been sleeping on the graphical forefront while Nintendo will be the one pushing that boundary by themselves, sure because Nintendo games are the ones that would mostly benefit from it with all their push for fidelity and photorealism right?

You have that backwards. 

It is going to be EASIER for Nintendo to get on par with this generation than the PS4/XB1 generation.

People can spin this all day long but the fact is this generation of machines have a much lower floor than the PS4/XB1 did. 

4 TFLOPS is much, much easier for Nintendo to get games from for 2023 than matching a PS4/XB1 was in 2015 for the Tegra X1. 

The equivalent of this is more like if Sony also released a PS4 Lite or something in 2013 that was only 1/3 the PS4 ... the Switch that can already handle ports of Witcher 3 and Dragon Quest XI and DOOM OK .... imagine for a moment many, many PS4 multiplatform games all had performance modes that operated at 1/3 the power of a PS4 ... anyone trying to claim that wouldn't be a major difference is a fucking liar. 

Again so many people are threatened by this idea lol, but whatever. The fact is the floor for this next gen is definitely lower because of the Series S. 

There's going to be a lot of games sitting on XBox Series S that are going to be much, much easier to port to a Switch 2 than the current dynamic that exists between Switch to PS4 to XB1. The Switch 2 doesn't need to exactly match the Series S either, if just needs to be in a somewhat similar ball park, even 2.5 TFLOP docked with a modern architecture like Ampere would do the trick for many, many games. 

A game running at 1080p on Series S (4K on Series X), but you have to understand because of DLSS 2.0 on Switch 2 that same game could run at 360p undocked and 576p docked resolution (upscales nicely to 900p undocked and 1440p docked). There what's left of your performance gap shrinks considerably. 

You don't need even 4 TFLOPS when you're only asking for a modern GPU design to display a resolution of 360p-576p (sub GameCube/PS2 level for undocked, lol).

Don`t forget the Unreal Engine 5 demo on PS5 at 1440p that Digital Foundry said they couldn`t really distinguish the pixels and that was basically making pixel count rather obsolete.

Also both MS and Sony have talked about half precision integer capability on their console that would support the temporary reconstruction technique.

Not to forget as well that PS4Pro made big use of single precision for their checkerbordered technique that since that time had DF saying their checkerbording from 1440p to 4k had almost the same level of clarity as native 4k for a fraction of the cost.

It is silly to think that everyone have been sleeping on the graphical forefront while Nintendo will be the one pushing that boundary by themselves, sure because Nintendo games are the ones that would mostly benefit from it with all their push for fidelity and photorealism right?

That's great and all, but I would say again ... where is it? They've shown really nothing of this on any actual XBox, which begs a lot of questions. 

I'm guessing their implementation of this that isn't non-hardware specific has drawbacks in performance cost. Otherwise they would be crowing about it from every roof top. 

Especially if Sony doesn't have an equivalent. Nvidia's DLSS 2.0 is not some smoke and mirrors PR buzz, you can run it on actual games right now. Series X is three months from launch and Microsoft has little to nothing to say about DLSS like implementation. That is pretty hard to believe given the performance implications something like that has. 

Unless of course it doesn't work as well in real world scenarios (or at least relevant to the AMD hardware) as MS has been saying. 

Well there is a slight catch as of right now based on Xbox Series X RDNA2 specifications.

Tensor cores are very specialized cores that can accelerate INT-8 TOPS which is what's used to Render DLSS according to Digital Foundry. The thing is, the Series X and Ps5 and probably the Series S will have cores that can also do them however not as fast as Nvidia's Tensor cores. Now due to architectural differences and software differences and etc, maybe RDNA 2 might not need them to be as fast so who knows. I am not that technical into how DLSS works at a low level loll.

The point is though, with what we know now, if we were to port DLSS over to the Series X. It would take twice as long to Render DLSS compared to a 2060.

And there's a bit of an interesting article here too about it:

https://www.eurogamer.net/articles/digitalfoundry-2020-image-reconstruction-death-stranding-face-off

"There's an important point of differentiation between Nvidia's hardware and AMD's, however. The green team is deeply invested in AI acceleration across its entire business and it's investing significantly in die-space on the processor for dedicated AI tasks. AMD has not shared its plans for machine learning support with RDNA 2, and there is some confusion about its implementation in the next-gen consoles. Microsoft has confirmed support for accelerated INT4/INT8 processing for Xbox Series X (for the record, DLSS uses INT8) but Sony has not confirmed ML support for PlayStation 5 nor a clutch of other RDNA 2 features that are present for the next generation Xbox and in PC via DirectX 12 Ultimate support on upcoming AMD products.

Broadly speaking then, the Xbox Series X GPU has around 50 per cent of the RTX 2060's machine learning processing power. A notional DLSS port would see AI upscaling take 5ms to complete, rather than a 2060's circa 2.5ms. That's heavy, but still nowhere near as expensive as generating a full 4K image - and that's assuming that Microsoft isn't working on its own machine learning upscaling solution better suited to console development (spoilers: it is - or at least it was a few years back). In the meantime though, DLSS is the most exciting tech of its type - we're sure to see the technology evolve and for Nvidia to leverage a key hardware/software advantage. The only barrier I can see is its status as a proprietary technology requiring bespoke integration. DLSS only works as long as developers add it to their games, after all.

As exciting as the prospects for machine learning upscaling are, I also expect to see continued development of existing non-ML reconstruction techniques for the next-gen machines - Insomniac's temporal injection technique (as seen in Ratchet and Clank and Marvel's Spider-Man) is tremendous and I'm fascinated to see how this could evolve given access to the PS5's additional horsepower."

With that being said though, there are a lot more months for both Sony, MS and AMD to reveal more and more about their GPUs and other features so things could change.

It's not really Nintendo that should push the boundary of graphics. It's more a question of Amd vs Nvidia technology. I believe a next-gen Switch will have its own advantages compared even to next gen PS and Xbox, just for the eventual use of latest AI technologies developed by Nvidia. But that said, Switch 2 (if it retains the current form factor) will still be a low power / low spec machine so I agree with you, people should not get their hopes too high on the graphical front. 

It's not really Nintendo that should push the boundary of graphics. It's more a question of Amd vs Nvidia technology. I believe a next-gen Switch will have its own advantages compared even to next gen PS and Xbox, just for the eventual use of latest AI technologies developed by Nvidia. But that said, Switch 2 (if it retains the current form factor) will still be a low power / low spec machine so I agree with you, people should not get their hopes too high on the graphical front.