Only two being sold to consumers.
Do you think there will be a mid gen refresh console from Sony and Microsoft?
- Thread starter invictis
- Start date
That's not the issue though imo. It's having 3 dev target specs.
arandomguy
Veteran
Those machines are simpler because they're using a completely off the shelf APU (well except Valve, but that's a longer story) and as such the large fixed cost for that is spread over all purchasers (mostly the broader consumer market). You can't really do that with the consoles as there is no off the shelf APU available. An XSX Pro would require a new APU, that's a much more complex undertaking than what those companies have to do for their products.
As an aside I feel that MS is going to have a challenge on the hardware side against Sony as long as they share a common supplier and the market dynamics continue with Sony being the larger and likely preferred customer. It's the second generation now and if you look at the overall aspect (this would include cost and complexity for the performance you get) I feel it's hard to argue that MS is not getting the worse end of the APU side.
What counts as a visibly superior experience? From whose point of view? From the perspective of the technocrat who overestimates every little difference, or from the perspective of the 90% of average users who see no significant difference between a 3070 PC VGA and an Xbox SeriesX or PS5 graphics?
In this generation, the console is the closest to the high-end PC experience, and this can easily be expected three years after their release. Most multiplatform games run on console with high PC settings, many at a stable 60fps. This could not be said in the previous generation after all this time.
So what a superior experience is is a matter of perspective.
If it's functionally the same hardware, only a bit faster, you can target the old base units and the faster hardware will just run it better. In theory, of course, so long as something that's broken in your game isn't exposed by higher performance.
The Xbox One S was an Xbox One with a slightly faster GPU and higher embedded memory bandwidth. Old games just worked on it, and bottlenecks allowing games ran a bit faster. It was never a 3rd spec that developers needed to 'target' as such. IIRC even PS4 Pro could run base PS4 games faster without the Pro needing to be targetted (using higher clocks, not using the full width of the butterfly GPU).
Series S and Series X would benefit from a shrink, a base clock bump, and some kind of boost system. Series X could also possibly benefit from going to a faster and narrower memory bus now that faster GDDR6 speeds are more plentiful and affordable.
That's a subjective question so not something I can definitely answer. We're looking at roughly the difference between a 2080/S and a 2080Ti in raster, so 15-20% which isn't huge but could for example be the difference between a sub 60 or locked 60 experience, or the difference between 4x AF and 16x AF plus other higher quality effects. But in RT we're looking at more like a 40-50% delta which could be the difference between RT on and RT off for example, or something akin to the visual differences between the consoles quality and performance modes. I know not everyone will notice those differences, but not everyone will notice the difference between the XBSX and XBSS, or the PS5 and PS5 Pro for that matter. There are however enough people that care about those kinds of incremental improvements to arguable justify a higher tier console. Afterall, 13m+ PC gamers own GPU's of 3070 calibre or above when they could have just settled for a 6650XT.
Even if you were to argue that what I describe above isn't a noticeably superior experience, could you say the same of the 3080 and above for example? I think most would argue a 3080 with a 50% raster uplift and roughly 100% RT uplift can certainly produce results that would be sufficiently superior to be noticeable to the average gamer (even if we ignore DLSS which should absolutely be considered if we are comparing user experience rather than raw performance). That could literally be the difference between 30 and 60fps, and that still accounts for around 6% of the PC userbase or around 8m users today. And it's worth noting that this new console isn't supposed to launch until late next year so the landscape on the PC side should be a fair bit different by then. That 6% today could comfortably be 10% by then, or again, more units than there are XBSX consoles for example. Half that number might be in the double and more raster camp too. That's is arguably enough to justify the need for a higher performance console.
I don't know why people bringing up Xbox One S...it was a slim revision that had marginally different clock speeds...not a pro console. The pro consoles were the PS4 Pro and Xbox One X
They are just pointing out that tweaking speeds shouldn't require a lot of work to get the game running on the system
No one had to do work on the Xbox one s despite having different clock speeds because the xbox one games ran on it just very marginally faster.
The ps4 pro and xbox one x had more work done to them but the majority of software just ran on them again.
The questions on the sony side is do they stick with zen2/rdna 2 with just more cu's and different clock speeds or do they go to newer technology like zen4/5 and rdna3/4. If they do go to rdna 4 and its a large enough difference is sony's dev tools robust enough account for what could be large differences in the capabilities of the machines.
I think fundamentally if sony were to go with say a zen5 + rdna 4 you could be looking at a large gulf than exists between a series s and x and we all know how people feel about the s. Not only could you be looking at double the performance or more but zen5 and rdna 4 could support features that doesn't exist in the ps5. That would be a much more difficult situation to over come than two systems at different performance levels of the same hardware zen2/rdna2
What's more if one company launches a zen2/rdna2 but more console in 2024 what would be the ramifications if another company decides to make the move to zen5/rdna 4.
Would Sony be able to over come having a brand new pro system launch in 2024 with zen2/rdna 2 while Ms launches a new system in 2025 with zen5/rdna 4 ? What would that even look like for sony? Would they have to limp along another 2-3 years with the pro console before releasing a new one , would they have to abandon the pro console right away and launch their own new system ? More so for Ms with its two prong approach to consoles what if they design a zen5/rdna4 machine that is similar in performance to the ps5 pro as the s console and creates a premium sku that is even more powerful with more ram and so on ? You can reverse the roles in this too obviously it would be doom for Ms if they launched an xbox series console with just more zen2/rdna 2 and sony put out a zen5/rdna 4
Of course rdna 4 could be just as bad as rdna 3 and all of this is moot and we are looking at another generation of AMD screwing up.
To me I rather both MS/Sony come out with completely new consoles sooner vs later. If we got completely new consoles in 2025 perhaps it would take until 2027 for them to be taken advantage of in more than just higher resolution and frame rate machines but at the same time we wouldn't be stuck with such poor gaming performance.
I concur. They should just name their name their next devices.
Xbox. Xbox Huge. Xbox to go.
Frenetic Pony
Veteran
There'd also be 0 point in doing just Zen/RDNA2 but "bigger". It'd be a total waste of money in both development and manufacturing for marginal at best compatibility improvements, as advancing gpu/cpu architecture would be giving you better performance nigh for free other than dev time. Which, as you pointed out, Sony would be asking devs to support another SKU. But they'd be doing that either way, that's a major headache no matter the arch.
Much better shove an overclocked GPU, new video encoder, new or even just overclocked CPU, and maybe if really needed 2gb LPDDR5 in for a streaming buffer increase. Potentially, as it might look the same to devs, split the memory bus off into chiplets like RDNA3 to lower manufacturing costs. Now you can stream games in 4k60hdr, and/or decode 8k AV1, and games hit more stable framerates, and the console is smaller. PS5s (too many S's?)
I mean the pro consoles last time were the same Jaguar CPU's with a higher clock...and a bigger GPU.
I don't see why they wouldn't follow a similar thing this time. The GPU is what's going to give you a PS5 game but a better resolution, fps, raytracing. Which is almost certainly what the selling point will be.
I don't see why they wouldn't follow a similar thing this time. The GPU is what's going to give you a PS5 game but a better resolution, fps, raytracing. Which is almost certainly what the selling point will be.
I still don't know if it's a good idea. They may as well bake forwards and backwards compatibility into the hardware or software from here on out if we get a Pro system. If they can get a 3x bump in ray tracing performance it may be worth it. Wouldn't mind seeing the the PS5 end up being a 1080p machine if it means we finally get next gen software to go with it. I guess it depends on how transformative the experience is with the games while still giving the base platform a good experience.
I do believe the market can bare an $600 price tag for the machine over the first year. If supply is very limited they could probably get away with an initial $800 price tag.
MS can just iterate a new series console at the high end to have parity with any ps5 pro. Even if they sey they have zero intention of releasing one the likelihood of them at least looking at the option is high. Yeah it's 3 sku's they'd need to target but that may be beneficial when it comes to development overall when considering PC versions. 3 different performance profiles while more work may allow them to have better focus on minimizing issues on the PC side. But yeah in this scenario I see the series s essentially becoming a 720p30 machine. Might not the the best idea for MS.
I do believe the market can bare an $600 price tag for the machine over the first year. If supply is very limited they could probably get away with an initial $800 price tag.
MS can just iterate a new series console at the high end to have parity with any ps5 pro. Even if they sey they have zero intention of releasing one the likelihood of them at least looking at the option is high. Yeah it's 3 sku's they'd need to target but that may be beneficial when it comes to development overall when considering PC versions. 3 different performance profiles while more work may allow them to have better focus on minimizing issues on the PC side. But yeah in this scenario I see the series s essentially becoming a 720p30 machine. Might not the the best idea for MS.
This is exactly what I would be afraid of. Take the ps5 and doubling the rdna2 cu's to 72 or something is going to have it still below even rdna 2 pc parts from 2020 let alone anything from amd in 2024 or nvidia in 2024. It be largely the same thing with RDNA 3 to be frank
Inuhanyou
Veteran
To be honest we shouldent really care about that. It'll still be better performance. If people want to splurge on a partial upgrade it should be up to them. It's called a half step for a reason. I don't see the point of them but clearly they have some sort of audience. As long as Sony only sells a fraction of their PS5 consoles as pro model and properly focus on reducing footprint and power draw and size with the slim as the much bigger priority let em sell the pro as a side thing.
This is partially my cope with the pro coming out whether I like it or not. But what are ya gonna do eh
I mean honestly not much other than marketing!
But "next generation console" still has significant marketing power I suspect.
When looking at actual HW or more specifically possible HW/SW features that could be enabled on a "True next-gen Console"..
Native high quality upscaling built into the APU, Frame-gen tech built into the APU + SDK from day 1.
more powerful RT features.
What is NV/AMD working on right now?
whats the current ground breaking tech in Graphics?
But honestly MS's smart play would be - imho, and i'm not expert,
Focus on Xbox as a platform, and all the games they offer at best possible quality.
Were in an era of diminishing returns for visible improvements to games,
adding 2x as many TFLOPS, aint gonna show on the screen for many,
but more quantity and quality might be a meaningful metric?
anyway, I'm still interested to see a PS5 Pro, but i'm not convinced it will be a commercial success
I think console gamers should want the best performance they can get at that price point. LIke I said in a previous post we have games like FF16 that run at 720p 30fps. Should gamers be happy if a ps5 pro or xbox seres xx gets them to 720p 60fps ?
Console gamers should be pushing for more comprehensive hardware upgrades esp since you have to throw the baby out with the bath water unlike with pcs. Sure zen2 may be enough on the cpu side to eek out a run to 2028 but if we get zen2/rdna2 again in consoles that wont be great for gamers in 2026. With a pc you can keep that zen 2 and jump from a radeon 6600xt to a radeon 8600xt or 9600xt in that time frame. consoles its the whole ball game again.
arandomguy
Veteran
Regarding the commercial success of Pro consoles what likely needs to be considered is not just the sell rate of the Pro hardware itself but attach rate of software and other ancillary revenue. If I had to guess the likelihood is that Pro console buyers likely skew higher into the spending range and that Sony (and maybe MS) possibly has data that might support a Pro console also driving other sales.
I thought that was the issue Iroboto was concerned with, consumer overload. For devs it's no worse than Pro or the 4 SKus of previous or targeting PCs. Perhaps even easier, if XBSX+ is XBSX at solid 60 fps with better RT, the kind of upgrades being talked about here. The plus would just run XBSX better without being a particular new target (which is what the Pros were).
Similar threads
