Ahh that makes sense. That was definitely the case with the transition, they are being much more bullish now about the node cadence but then again ASML have great incentive to say that!
-
Ever wanted an RSS feed of all your favorite gaming news sites? Go check out our new Gaming Headlines feed! Read more about it here.
-
We have made minor adjustments to how the search bar works on ResetEra. You can read about the changes here.
- Status
Ahh that makes sense. That was definitely the case with the transition, they are being much more bullish now about the node cadence but then again ASML have great incentive to say that!
if its just ps now, i really have no interest. hope i can buy or download my digital copy of ps1/2/3 game on ps5 and play local.
Clocking the GPU high makes everything in the GPU faster. Its not rocket science, this speaks nothing of architectural differences. Even though GCN was inefficient when it comes to gaming it still performed high when it comes to just compute workload. The major change in RDNA that produced the greatest performance was the redesigned wavefront which made it easier for all queues to be easily utilized more efficiently compared to GCN. RDNA uses a narrower wavefront with 32 work items which is much more efficient and simple than GCN older wave64 design and this gave them an almost 2X IPC uplift. It is along the same vein Mark Cerny gave reasons why fewer higher clocked CUs being able to be fully utilized more effectively than more CUs clocked lower. If clocking higher doesn't provide any benefits then AMD and Nvidia will still be clocking their GPU at 1GHz and just add more compute units, instead we see a steady rise in GPU clock speed, and PS5 being one of the highest clocked at 2.23GHz. XSX is not clocked low either, it is 1.825GHz coming from last gen that was clocked at sub or near 1GHz.
Last edited:
Is it not like we've been saying for years... PS5 tech will end up being installed in data centres for future PS Now. It would then of course be ideal if PS5 could then replace the PS3s they're currently using by emulating. And we all know PS5 is more than powerful enough to emulate PS2, PS1 was being emulated even on Vita ffs... Maybe this is finally confirmation that yes, PS5 can emulate all previous consoles, and that will also be replicated in their future PS Now data centres. Bring it on if true.
j-platpat
www.j-platpat.inpit.go.jp
Patent for Sony cloud system, incluind full BC, there is some stuff about streaming but the patent is huge.
If I'm not mistaken, PS3 games are run off actual PS3s server blades in their server centres. This suggests they'll now be run off PS5s emulating PS3s. That definitely gives them the option to bring that emulation to actual PS5s right?
Yah, that is what the patent is saying.
Their is also a lot of stuff about share gameplay, social media and the ability to load games from peopel videoclips.
even if it was filed back then sound like this is a revision to it so there are probally additions to this and possibly now they are realizing what they wanted to do back then.
Depends.
I don't have time to look into this, but you (generally speaking) can't add new disclosure to a patent application that's already been filed.
However, this could be a continuation-in-part that claims priority to an original filing back in 2013. If that's the case, you certainly can add new disclosure.
Thanks, I had missed this. Definitely not a unified L3.I don't believe at the 8 cores CCX theory for PS5 CPU. You remember Flute benchmark ? It clearly showed 4MB L3 by CCX and I doubt they are going to use 4MB in total, it's going to be 8MB in total using 2 CCXs. But I think Sony could have customized their CPU in order to reduce caches latencies which is going to be very important for PSVR2.
Always check the original filing date.
We don't always agree on many things, but this, yeah, I totally agree with this comment. I think we will see Slimmer designs, but not a refresh.
Lol well you can.
We never needed a resolution bump that justified a mid gen refresh before the Pro/One X. Up till that point, Sony and MS were refining their consoles with mid gen refreshes (MS added Wifi, built in memory rather than a memory card in their Xbox 360S, for example) ,Sony were trying their hardest to lower cost with PS3. But even if we go all the way back to the PS1 Slim.. it was purely for form factor and the PS2, from what I recall, did get a tiny performance boost, but I can't remember why it was done (early models with 294MHZ and later models were 299MHZ which was basically the same)
So yes, logically, I think you're on the money on that comment. I'm sure we will have some sort of upscale hype this gen to support 8K tv's.. but nothing touting 8K Native. We're back to expecting Mid gen refreshes purely for Quality of life benefits.
How much of the mid-gen update was influenced by being able to experiment on new technology /apis / backwards compat / other stuff before a whole new gen arrived and things were set in stone for another 7 years?
that's part of the reason I personally think midgen might happen again, but maybe it's not actually all that influential/important?
You don't need new hardware (in terms of the main processors) for that.
I could see new revisions with the latest WiFi tech for example. If a newer HDMI standard launches and is relevant to something like PSVR2.
Otherwise, I don't see faster hardware that delivers more consistent native 4K (versus reconstructed 4K that 90% of gamers won't be able to discern any meaningful difference between) to be a strong enough business case for spending the additional billions in engineering a new mid-gen platform.
For some things you do - RPM, Checkerboard ID Buffer, Delta Color Compression, etc. https://www.eurogamer.net/articles/...tation-4-pro-how-sony-made-a-4k-games-machine
None of those features alone are sufficient to justify making new hardware.
The business case for the Pro and X1X was 4k gaming. All the other additions were either in service to the this (e.g. ID buffer) or just sundries added because of the newer GPU arch.
Unless you're part of the business, you don't know that. Which is why I asked Albert if he could provide his insight on how valuable that sort of thing is.
All I'm trying to do is to gain insight into what's important to console manufacturers and devs :)
The kind of consumer this is targeting would be motivated to upgrade for things like continued ultra settings, higher base resolution, higher framerate, and high ray tracing settings.
I think MS's memory usage experiment on the X1X was a pretty interesting way to benefit next-gen using the mid-gen refresh. The diskless X1S also seemed like a way to test how the audience will respond to a diskless system. I also think both mid-gen console laid the ground for developers having to support more than one SKU at the same time and the idea of a more powerful console that is BC but also can receive an upgrade patch. Developers probably feel much more comfortable with that today then they did in 2016. And I'm sure both the Pro and X gave MS and Sony some insights into BC, especially how the Pro is doing BC on the hardware level with a boost mode and the PS5 seems to be doing the same.
Last edited:
You're describing a PC gamer. Consoles don't target PC gamers.
His point and opinion on the necessity and feasibility are valid, as are others who have already pointed out the path for hardware and node transitions, chiplets, etc. I was expressing that an increase in graphics fidelity, rasterized or ray traced is the kind of thing that would motivate this sort of purchase. Many on ERA have even mentioned the willingness to upgrade if they could get 60fps on every title for example.
Consoles target people playing on PC just as much as customers from more directly competing platforms. The same reason why the Pro was meant to keep and expand the amount of enthusiasts in the ecosystem. Increased fidelity and performance more akin to upgrading a PC, but with the convenience of a console is not that hard of a sell.
yeah agreed - software development has certainly embraced the "release smaller things more frequently so you can actually know what customers love or not, instead of building a product in the dark for a long time and then releasing it, only to find out you built the wrong thing." mentality.
"We're agile!" anyone? :P
and that's where I feel like the midgen releases allowed the hardware teams to do a little bit of the same - get early feedback on new features, see how it works (or doesn't work), and then you have good data for the next full-gen release to know quite a bit better about what actually works and what doesn't work and what didn't work but could if these changes were made type of thing.
How likely is it that Sony is still hiding full BC, what do you think? How different would it be to pull that actually off? Kinda suspicious they didn't show the feature to have BC with PS4 games on their last PS5 event, which also aimed towards casual gamers that might not know about this. Maybe there are some more things that we don't know about yet.
I would actually prefer to have PS6 for Holiday 2026, with no midgen PS5 Pro.
However if this is going to be another very long generation like last gen was (7-8 years) and NEXT gen won't happen until 2027 or 2028, then I'd probably be on board for for mid consoles in 2023/2024.
Ideally though, the base PS5 is powerful enough in terms of CPU power that it won't struggle as badly later in its life as the 2013 consoles did with Jaguar.
However if this is going to be another very long generation like last gen was (7-8 years) and NEXT gen won't happen until 2027 or 2028, then I'd probably be on board for for mid consoles in 2023/2024.
Ideally though, the base PS5 is powerful enough in terms of CPU power that it won't struggle as badly later in its life as the 2013 consoles did with Jaguar.
I will point out that this forthcoming generation of consoles is likely to be even lower margin than last gen (not just cost, but also power and thermal as these designs are much more optimized to the edges), and the additive cost of more transistors is no longer going down as node processes evolve. It will take the entire span of a generation to have a meaningful increase in performance for a given price, particularly because of the diminishing returns of further fidelity increases.
The June event didn't really show OS, UI, or hardware features, it was more game-centric. I'm sure that we will see BC games once they unveil the OS and its' features.
Not disagreeing with this and at the end of the day, the possibility of these mid-gen systems are years away. Simply reaffirming that the market for these enthusiast upgrades are sold on enthusiast features, and they do not necessarily need a 4k buzzword for it to be successful.
No that's describing every console gamer (like me) up already had a base PS4 or XB1 and jumped at the chance for a new mid-gen refresh when they launched. Console gamers like the simple life sure, but can still want the best experience possible.
Is this something new?
www.ign.com
Sony Patent Suggests PS1, PS2, and PS3 Could Be Backwards Compatible Via the Cloud - IGN
The patent suggests older PlayStation consoles could be emulated via a virtual machine that mimics their operating systems.
Sony and MS will release next gen mid gen consoles for the following reasons:
- Keep players invested in their eco system. Dont want anyone going to PC or worse buying a competitor's system.
- Sell profitable consoles while cutting prices on base consoles selling them at cost or for a loss.
- Offer gamers better ray tracing and 60 fps games.
- Keep revenues from declining like they do in the second half of each console generation.
Price wont be a factor. Gamers who buy mid gen refreshes dont care about money. If they are willing to spend $1500 building a PC, they are willing to spend $500-600 on a mid gen refresh. Less, if they trade in their existing console like we all did.
- Keep players invested in their eco system. Dont want anyone going to PC or worse buying a competitor's system.
- Sell profitable consoles while cutting prices on base consoles selling them at cost or for a loss.
- Offer gamers better ray tracing and 60 fps games.
- Keep revenues from declining like they do in the second half of each console generation.
Price wont be a factor. Gamers who buy mid gen refreshes dont care about money. If they are willing to spend $1500 building a PC, they are willing to spend $500-600 on a mid gen refresh. Less, if they trade in their existing console like we all did.
I think you are pretty much right. It's honestly already hard to tell the difference between the PS4 and the PS4 Pro for a lot of titles, and that's an over 100% improvement in specs.
I think they are already cramming the maximum amount of tech/wattage/heat into a console sized box with these next-gen consoles that they don't really have any drawbacks or caveats to be eliminated or addressed any time soon. They can only potentially get better in proportion to how much more efficient per-watt CPU/GPUs get in 4 years time. I would guess maybe 50% faster, at most. Not noticeable and probably not worth it. 2TB slimmer revisions on the other hand...
GPU, CPU, SSD are all pretty much 'maxed out' for something that fits in a vaguely console sized box in 2020.
I just don't see the "move to PC" being alluring this time around. SSD, High-frame-Rate, and overall performance advantages are pretty much at minimum nullified by these consoles, and at a much lower cost.
I don't see this changing by 2023 or so. Or at least changing enough that it's worth re-engineering to prevent ecosystem loss.
I can hardly imagine the games we will be playing in 2028/2029, with consoles that launch Holiday 2027/28 and use mid decade technology that is probably barely visible or not visible at all on today's silicon/RAM/tech public roadmaps, aside from projecting process nodes down below 3nm.
I think you are underestimating what changes would be appealing to this market. The Pro primarily increased resolution with little effect on settings and framerate, the opposite would be enough to entice buyers.
- Status