Do you think there will be a mid gen refresh console from Sony and Microsoft?

[Shifty Geezer]

I wasn't suggesting that a PS6 would be a 'whole generation' ahead. But 2-3 years is absolutely enough to have a very decisive technical advantage.

Then it's surely enough for MS to have a decided advantage too? If 2-3 years later is enough for PS6 to differentiate itself from XBSY, then 2-3 years (4+ when it comes out) should be enough for XBSY to stand out from PS5 as PS6 will then stand out from XBSY. In fact arguably even better, because progress is reducing over time. 3 years advantage between 2025-2028 will be equivalent to 2.5 years advantage between 2020 and 2022.5

 

[function]

Nvidia are a good 3 - 4 years ahead of AMD in terms of RT performance, and arguably even further ahead in terms of upscaling.

Now this may change of course, but it's at least possible to go 2 - 3 years earlier and end up with something overall better.

This isn't to say it's realistic that MS could do this, but adding sufficient technology to AMD IP such as a powerful NPU or tensor cores, and making software to take advantage of that (e.g. DLSS), might mitigate some of the deficits of going earlier. It would be expensive though, and there's the question of whether there would be enough time.

Of course, Sony could do that same thing.

 

[Nisaaru]

I

Nvidia are a good 3 - 4 years ahead of AMD in terms of RT performance, and arguably even further ahead in terms of upscaling.

Now this may change of course, but it's at least possible to go 2 - 3 years earlier and end up with something overall better.

This isn't to say it's realistic that MS could do this, but adding sufficient technology to AMD IP such as a powerful NPU or tensor cores, and making software to take advantage of that (e.g. DLSS), might mitigate some of the deficits of going earlier. It would be expensive though, and there's the question of whether there would be enough time.

Of course, Sony could do that same thing.

If you look at AMD's GPU designs they tried to play smart by reusing existing GPU blocks which indicates that they squeezed RT into their limited transistor budget than throwing the kitchen sink at it.

I don't really think RT is about a 3-4 years engineering advantage but more about transistor budget, TDP/heat and "judgement" how relevant the function was for games at the target window.

If you look at the history of failing 4090(Youtube is full of videos covering this) due to all kinds of power issues NV's product strategy isn't without risk either.

*If* AMD's RDNA3 GPU design didn't fail so badly about TDP/clock they might have had at least a very competitive priced product vs. <=4080 in the last generation. This screwup "seems" to have bleed into previous plans for RDNA4 too.

 

[Frenetic Pony]

Nvidia are a good 3 - 4 years ahead of AMD in terms of RT performance, and arguably even further ahead in terms of upscaling.

Now this may change of course, but it's at least possible to go 2 - 3 years earlier and end up with something overall better.

This isn't to say it's realistic that MS could do this, but adding sufficient technology to AMD IP such as a powerful NPU or tensor cores, and making software to take advantage of that (e.g. DLSS), might mitigate some of the deficits of going earlier. It would be expensive though, and there's the question of whether there would be enough time.

Of course, Sony could do that same thing.

It's just ray traversal, nothing special. Intel did it, Apple did it and better than Nvidia with M3, AMD has had trouble at its RTG division for years now and finally fired their chronically underperforming division boss months ago. Maybe now they'll get to do what the rest of AMD manages, which is produce great hardware. MI300 is technically far more advanced than Nvidia's Hopper series or the upcoming Hopper Grace and has been shipping to HPC since last year, they can keep up with and surpass Nvidia as long as they can work around Nvidia's monpolistic play at buying up all the manufacturing.

Nvidia isn't special, they're often chronically behind in overall hardware efficiency but have super aggressive software, manufacturing, and PR to make up for that. Which is great for gaming, gamurs will buy PR hook line and sinker a good 80% of the time, margins are tight but large contracts help there if you're already established, and software (cough DLSS etc.) has its advantages. It's somewhat less effective in AI. The advantages they have there is they were the only game in town for a good while and software. But one company can't beat literally the entire world when it comes to software, MS just gave up against Linux in a lot of ways eventually and the same will happen here. PR is nonexistent, oh Nvidia keeps trying and gets morons to buy into it, but morons aren't in charge of spending hundreds of millions to billions on hardware contracts, none of their ridiculous PR is going to imping some unimpressed engineer at a cloud provider.

It'll be funny to watch Nvidia fail miserably on CPU. The entire market is already way overly competitive, there's 0 advantage in software, basically none in PR, and good luck on the manufacturing front when margins are shrinking to non existence while its massive competitors with established and stable foundry relations, or their own foundries (cough Intel) can beat it outright there. Jen-Hsuns playbook just isn't going to work there.

 

[Kaotik]

If the pro soc is not going to have ada size caches and features like SER just adding bvh traversal hardware would be closer to ampere right? Same goes for the AI hardware, not sure expecting to match nvidias current gen tensor cores on AMDs first attempt is logical but I don't think matching these is as important, adas tensor cores are probably overkill for just upscaling.

First attempt? AMD has already 2 gens of matrix crunchers (what you called "ai hardware", exact same thing as tensor cores or Intel's XMX units) for GPUs as well as 1st gen of NPUs (XDNA)

Nvidia are a good 3 - 4 years ahead of AMD in terms of RT performance, and arguably even further ahead in terms of upscaling.

Now this may change of course, but it's at least possible to go 2 - 3 years earlier and end up with something overall better.

Err, maybe you should actually look some benchmarks (not just single cherrypick) next time

This isn't to say it's realistic that MS could do this, but adding sufficient technology to AMD IP such as a powerful NPU or tensor cores, and making software to take advantage of that (e.g. DLSS), might mitigate some of the deficits of going earlier. It would be expensive though, and there's the question of whether there would be enough time.

Of course, Sony could do that same thing.

Why would you add something like that to AMD IP when AMD already has them in their IP pool?

 

[t0mb3rt]

I


If you look at AMD's GPU designs they tried to play smart by reusing existing GPU blocks which indicates that they squeezed RT into their limited transistor budget than throwing the kitchen sink at it.

I don't really think RT is about a 3-4 years engineering advantage but more about transistor budget, TDP/heat and "judgement" how relevant the function was for games at the target window.

If you look at the history of failing 4090(Youtube is full of videos covering this) due to all kinds of power issues NV's product strategy isn't without risk either.

*If* AMD's RDNA3 GPU design didn't fail so badly about TDP/clock they might have had at least a very competitive priced product vs. <=4080 in the last generation. This screwup "seems" to have bleed into previous plans for RDNA4 too.

This. The difference between nVidia and AMD in terms of ray tracing and upscaling isn't a time thing. It's just an implementation thing. They chose different solutions to the same problems. There's no reason AMD couldn't catch up performance wise with their next gen GPUs if they choose to spend money and transistor budget on developing those things and subsequently price their GPUs higher.

 

[Shifty Geezer]

Any more nVidia versus AMD and I'll nuke posts. Keep this thread on topic.

 

[vjPiedPiper]

Theres also a dedicated thread for Next-gen console predictions too.

As a XSX owner i would be OK with MS, moving the standard forward and declaring the current series X as the new - low-end system, and introducing a new high end.
But i'm not sure it's a good business move...

Of course plenty of games are sub 1080p native. on a xsx already. we probably need a bigger jump even for the low-end of next-gen.
I'm skeptical of how much visual improvement a PS5Pro will bring?
Other than DF style comparisons, are users really going to notice a native res of 900p vs 1400p?

 

[see colon]

Of course plenty of games are sub 1080p native. on a xsx already. we probably need a bigger jump even for the low-end of next-gen.
I'm skeptical of how much visual improvement a PS5Pro will bring?
Other than DF style comparisons, are users really going to notice a native res of 900p vs 1400p?

One of the things I noticed about Xbox One and PS4 vs their X and Pro variants the most was the decrease in load times. Assuming there won't be a large difference in load times this time around, you'd be left with image quality or framerate differences, which I think are much less noticeable to everyday people. I know it might sound crazy, but I bet there are plenty of PS5's and Series X's hooked up to non-4k displays, or cheap ones with lackluster panels that will hide some of those gains as well.

 

[Johnny Awesome]

MS did ok in NA this holiday, but they got murdered in Europe. Japan's always been a lost cause.

At this time anyone who bought a PS4 doesn't really have a reason to get an Xbox instead of a PS5. MS is just barely selling Xboxes to most of the Xbox One users. They aren't converting almost any PS people.

They might not need to given their strength on PC, but if I were Phil and going to make the attempt, I would put those 40 studios to work making truly impressive games and really try to wow people with a new box in 2026 and put the X in an S sized box and put the S in an M handheld.

 

[Seanspeed]

Then it's surely enough for MS to have a decided advantage too? If 2-3 years later is enough for PS6 to differentiate itself from XBSY, then 2-3 years (4+ when it comes out) should be enough for XBSY to stand out from PS5 as PS6 will then stand out from XBSY. In fact arguably even better, because progress is reducing over time. 3 years advantage between 2025-2028 will be equivalent to 2.5 years advantage between 2020 and 2022.5

Again, I feel we're talking in circles here. No, if MS have an advantage for 2-3 years by releasing their 'next gen' system early, nobody will care because it will have no exclusive games. It'll just be Xbox Series X Pro basically, and it'll come way too late for that kind of device(I also felt Xbox One X was a bit late, but it at least delivered on the brief of 4k gaming much better than PS4 Pro did). All while Playstation 6 will come along, stomp all over it, and actually get next gen support from developers, cuz absolutely nobody will be ready to move on before 2028.

Xbox is not popular enough to get away with this kind of thing. They dont have the pull to move the needle via hardware. And not enough people are gonna buy an expensive Xbox super late into the generation for games still ultimately being built for XSX/PS5 as a baseline. People will be talking about actual next gen by then.

I appreciate the 'devil's advocate' discussion here, but there's genuinely no way this works out in anything except disaster for Xbox. They will have to spend incredible amounts of resources for this 'next gen' device that wont actually have next gen games and wont even sell all that well, only for the far more anticipated Playstation 6 that will be technically superior to come along and deliver not just want gamers want, but a platform that developers will actually be fine moving to develop for properly.

EDIT: I could maybe imagine, in some unlikely fantasy world, a scenario in which Sony fumbles hard with their big name releases in the latter half of the generation, all while Xbox does the opposite and releases a slate of huge must-play critical darlings in the next couple years, changing the reputation of these brands a bit more towards like what it was in the early 360/PS3 days. All while Xbox announces that Elder Scrolls 6 will be available exclusively for their new next gen machine within like a year of release, and the game knocks it out of the park. MAYBE then, I can see a case for Xbox getting away with it, by sheer reputation of great exclusive games. But they've got a long way to go to gain that kind of reputation again and not very long to do it. Dont see it happening.

 

[Shifty Geezer]

They will have to spend incredible amounts of resources for this 'next gen' device that wont actually have next gen games and wont even sell all that well, only for the far more anticipated Playstation 6 that will be technically superior to come along and deliver not just want gamers want,

This is the issue I have. You say PS6 will release games that stomp all over XBSY - yet they won't, because they'll be marginally better as the tech doesn't support that level improvement. And devs won't develop specifically for it but will instead be making cross-gen games again. Even Sony will be making cross-gen and cross-plat games! There's no scenario AFAICS where PS6 comes out three years later than something launching from MS in 2024/2025 and blows it away.

 

[cheapchips]

Are MS thinking about launching a next gen machine first just so that they have a box that doesn't require COD graphics parity?

Said partly in jest, obvious, but a couple of years of "this is the best place to play popular third part gaas game" is not to be sneezed at.

 

[Seanspeed]

This is the issue I have. You say PS6 will release games that stomp all over XBSY - yet they won't, because they'll be marginally better as the tech doesn't support that level improvement. And devs won't develop specifically for it but will instead be making cross-gen games again. Even Sony will be making cross-gen and cross-plat games! There's no scenario AFAICS where PS6 comes out three years later than something launching from MS in 2024/2025 and blows it away.

When I say 'stomp all over', again, I'm not talking about some generational difference, I'm just talking a decisive technical advantage. One in which there will basically be a very clear winner in every single game, putting an undeniable negative mark on the Xbox in basically every discussion about what console to buy(this absolutely hurt the Xbox One). Especially since Playstation is already the platform that developers prioritize, which will be a double whammy for Xbox. And probably a triple whammy in that such a rushed 'next gen' Xbox will probably not sell well, leading devs to prioritize it even less than they do nowadays.

I just dont understand what the selling point of this early next gen Xbox is. Just to have a couple years late in this current gen with slightly superior multiplatform results? The long cross gen period this time around kind of worked out because both Xbox and Playstation dealt with it in the exact same timeframe as each other. If Playstation still has to go through the 'cross gen' period, then Xbox is gonna have to go along with it as well, cuz Playstation will be the lead platform and the one that devs will wait for to actually start building towards true next gen titles. Meaning Xbox would be looking at like 5 years of 'cross gen' before people who bought it early on will get actual next gen titles for it.

Not to mention that while Sony is doing more cross-gen nowadays obviously, they still did have at least some next gen titles earlier on. Unlike Xbox which had none. I think Microsoft would have to change and demand that such a new Xbox also get some actual exclusive 1st/2nd party titles to stand any kind of chance. But then that would probably also piss off Series X/S owners who, because of the long cross gen period they dealt with, would barely feel the generation had even started and are now getting abandoned. Not to mention that these next gen titles, should they even get made, would be made with likely lackluster next gen tools and not enough time, on a 'not really a big enough leap' piece of hardware.

I can probably keep listing more and more reasons this would fail the more I think about it. It would just be such a bad choice.

 

[Janne Kylliö]

Theres also a dedicated thread for Next-gen console predictions too.

As a XSX owner i would be OK with MS, moving the standard forward and declaring the current series X as the new - low-end system, and introducing a new high end.
But i'm not sure it's a good business move...

Of course plenty of games are sub 1080p native. on a xsx already. we probably need a bigger jump even for the low-end of next-gen.
I'm skeptical of how much visual improvement a PS5Pro will bring?
Other than DF style comparisons, are users really going to notice a native res of 900p vs 1400p?

Yeah. I've been disappointed how badly has this console generation aged. Or maybe it's just poor artefact ridden upscaling that makes IQ suffer and eyes bleed. 1080p -> 4k with DLSS would be OK.

MS needs to talk with Nvidia/Intel/Qualcomm when it comes to next gen systems if they want to make clearly better system than what PS5 Pro likely will be. They can't really gain real advantage over Sony tech wise, if both companies are using the same gen AMD tech, produced with the same the process to match a similar price point.

Qualcomm ARM cores + Nvidia GPU or Intel CPU + Celestial/Druid based GPU might make more sense. ML based upscaling and strong RT (possibly even enable path tracing in consoles) would be nice, and I believe people will be ready in 2026 for a true next gen. I know I will.

 

[see colon]

Again, I feel we're talking in circles here. No, if MS have an advantage for 2-3 years by releasing their 'next gen' system early, nobody will care because it will have no exclusive games.

Maybe. But Playstation hardware crushed Xbox hardware this year, and Playstation only released 3 games this year, 2 of which were for PSVR2, while Xbox+Bethesda released 11 games, 7 of which were Xbox console exclusives. Brand loyalty is a hard thing to tackle. I don't mean this in a console warrior way, but a more measured, logical way. If you've invested in any way in the hardware, or more importantly, the software, for a console, changing sides means starting that library over. Now that people have a substantial digital library, even if it's just games they got from PSN+, it's a value downgrade to forsake that value for a new piece of hardware to play games or franchises you have not relationship with yet. Looking at the future, I see Xbox has made waves with the Blade announcement at TGA, but that game is made by a studio that is famous for making good games that nobody buys.

 

[QPlayer]

I see that the topic is more and more on the agenda, it concerns many people. Even now, I can only say that there was an original Xbox for just over 20 million compared to a PS2 for 120 million. No one would have believed that this could be reversed, everyone was convinced that those who bought a PS until then would not buy an Xbox. And yet, they sold 80 million of the X360, and it was an unreliable machine. If the machine was stable from the beginning and there was no negative feedback, even 100 + million would have been sold.

Nothing is set in stone, with proper marketing and lots of exclusive games you can blow up the console market at any time, but that marketing has to be very good!

 

[QPlayer]

At the moment, there seems to be no point in launching more modern, more powerful hardware on the market, because it is unfeasible due to the costs of the current hardware parts at an acceptable price.

In place of MS, what I would do now is to re-release the X and S series next year, as stated in the leaked document. I would choose a time to throw in the New design for the new controller with 5-6 exclusive games at the same time, as if it were a new launch, all at a cheaper price of $400/$200. If the Model S is that important to them, then make sure it gets to the masses, even if they lose 1-2 billion on the hardware, to increase future Game Pass subscribers.

 

[MCharta]

I see that the topic is more and more on the agenda, it concerns many people. Even now, I can only say that there was an original Xbox for just over 20 million compared to a PS2 for 120 million. No one would have believed that this could be reversed, everyone was convinced that those who bought a PS until then would not buy an Xbox. And yet, they sold 80 million of the X360, and it was an unreliable machine. If the machine was stable from the beginning and there was no negative feedback, even 100 + million would have been sold.

Nothing is set in stone, with proper marketing and lots of exclusive games you can blow up the console market at any time, but that marketing has to be very good!

i agree with that at the same time this is me looking for XBOX strategy to build a strong brand image :

 

[Sonic]

I see that the topic is more and more on the agenda, it concerns many people. Even now, I can only say that there was an original Xbox for just over 20 million compared to a PS2 for 120 million. No one would have believed that this could be reversed, everyone was convinced that those who bought a PS until then would not buy an Xbox. And yet, they sold 80 million of the X360, and it was an unreliable machine. If the machine was stable from the beginning and there was no negative feedback, even 100 + million would have been sold.

Nothing is set in stone, with proper marketing and lots of exclusive games you can blow up the console market at any time, but that marketing has to be very good!

That was a time when Sony were at their worst and MS was at its best. If we could see MS return to doing their best that would be a great help to rectify perception of the Xbox brand. Phil Spencer being all smug and cringy doesn't help them.