I don't know if 16Gbit density GDDR5 is planned, but they'd still have to deal with the I/O width per DRAM chip.
Last edited:
Next Generation Hardware Speculation with a Technical Spin [2018]
- Thread starter Tkumpathenurple
- Start date
-
- Tags
- 5 bytes
- Status
Gamers will likely have built up a library of titles - perhaps a considerable one, worth hundreds of buckaroos or more. Assuming they don't instantly stop caring about all that just because they bought a PS5, what would be the alternatives?
A: keeping a PS4 around for playing their old games.
Means two consoles to sit on a shelf taking up space, two inputs occupied into the TV, separate set of peripherals (controllers mainly) to keep charged and add extra clutter in your life with, more cable mess on the floor, more stuff to dust off every once in a while, hoping PS4 doesn't break after production has stopped so that games will become inaccessible. Plus anything I may have missed.
B: simply play all your games on PS5.
I don't know about you, but I know which alternative I would choose...
That's seems like a pretty accurate prediction imo. Though i think it's possible the CPU has less cores and/or slower clocks.
$399 definitely hits a mass market in a way that $499 doesn't. Especially on a global basis.
One of the things i remember with MS taking the lead with BC. Is that they specifically mentioned that all their contracts old and new were being re-written with BC in mind. So that they wouldn't need to go through this song and dance as they did with 360 games again going into next gen.
I was thinking redesigned for gddr6, each 16bit channel at 12-14 seems to be very similar to a single gddr5 32bit at 6-7gbps. Practically the same timings, commands, and the doubled prefetch (at half width) end up with the same granularity. The PHY is double the speed, but commands and data throughput per controller would stay the same?
One question is whether lowering the chip count and bus width, and memory power, is a bigger gain over the small increase in cost with gddr6.
Also I haven't figured out what happens to the banks count, since each channel have it's own half of the chip's dram arrays.
All people speak about power and spec but I hope Sony will improve the build quality of PS5 because I had problem with the first serie of DualShock 4, the plastic looks cheap and the noise is there... They are in a much better financial situation than after the PS3, they need to step up...
Microsoft buid quality is much better....
Microsoft buid quality is much better....
Last edited:
D
Deleted member 86764
Guest
If you were to look at Microsoft's hardware from the designers' perspective* the Xbox 360 was quite powerful, but initially sold at a loss. Comparatively, the Xbox One was underpowered and I think sold at a profit (or let's say, less of a loss). The Xbox 360 was also much more of a success. So internally at Microsoft you'd think there would be an argument in favour of making something fairly exotic or powerful.
Sony on the other hand have had something of the reverse situation; the less exotic hardware being more successful and the simple(r) hardware being more successful.
If both manufacturers were to base the hardware design on their more successful consoles, it would suggest that the Xbox 9 (yeah, why not) would be much more powerful than the PS5.
*ignoring marketing completely - which may be a touch foolish considering how bloody awful the marketing for Xbox One was
Sony on the other hand have had something of the reverse situation; the less exotic hardware being more successful and the simple(r) hardware being more successful.
If both manufacturers were to base the hardware design on their more successful consoles, it would suggest that the Xbox 9 (yeah, why not) would be much more powerful than the PS5.
*ignoring marketing completely - which may be a touch foolish considering how bloody awful the marketing for Xbox One was
The mid gen refreshes are supposed to mature into something that can hold the fort while we wait for the next big architectural change. At least this is the way I've always seen them. Until they are at their 3rd year of operation, I don't really think we've seen them either running at full tilt yet. There is still hardware that can be extracted further, more learnings for developers to use as a stepping stone for some features for next gen.
D
Deleted member 11852
Guest
I agree, I think it's something they'll do as long as it's not a) time consuming, b) complicated, c) compromises anythng about PS5 they consider key. But Sony can't be so stupid not to be aware that if PS5 does not have, at least, compatibility with PS4 (even if not much in the way of enhancements), then PS4 owners cutting lose their software libraries are more much likely to jump to Xbox. Not definitely, but certainly more likely.
Tkumpathenurple
Veteran
If the PS5 doesn't possess BC, it will likely be reason enough for me to go from not owning an Xbox to making the XboxTwo my primary console. Then, a couple of years down the line, get a PS5 for its exclusives.
I read on Resetera that a patent was filed by Mark Cerny, after the release of the Pro, all to do with backwards compatibility, so hopefully it's in hand, as I'd prefer not to piss about with two consoles.
I do think it's important to stress that the PS4 is the first numbered PlayStation not to possess BC at all, so, this line I've read at a few places across the Internet, that Sony doesn't care about BC, is absolute nonsense.
I read on Resetera that a patent was filed by Mark Cerny, after the release of the Pro, all to do with backwards compatibility, so hopefully it's in hand, as I'd prefer not to piss about with two consoles.
I do think it's important to stress that the PS4 is the first numbered PlayStation not to possess BC at all, so, this line I've read at a few places across the Internet, that Sony doesn't care about BC, is absolute nonsense.
In the probabilities of things, the longer it takes for PS5 to release the higher the probability for BC. If it were released this year like SA somewhat implies i would give it a low probability, at least, to ship with BC.
I think Sony can implement any hardware customizations to enable BC and thus I expect it to be supported in a hardware form. The business of BC is something else entirely. Licensing agreements, or how licenses are handled are a different matter which are MS greatest hurdles at the moment.
It was a patent to change the timings of a piece of hardware test BC I believe.
D
Deleted member 13524
Guest
The only reason for the PS4 to not have BC is because it would require gigantic performance to emulate a a 240 GFLOPs CPU that ran 7 units heavily specialized for floating point performance at 3.2 GHz.
Had the PS3 CPU been used for exclusively for non-graphics tasks, then it might have been easier to emulate, but the devs ran pixel shaders on the Cell due to a GPU that was lackluster in pixel shading performance at the time.
So Sony could either go with embedding a 3.2GHz Cell co-processor (whose development had ended in 2009) into their SoC that would either make it a lot more expensive to make or take away precious GCN GPU resources, or they could have thrown more money at further evolving Cell (a path they opted out very early according to Mark Cerny) or they could drop BC and have the higher performing and cheapest console launching in 2013.
The PS4 doesn't have anything as exotic as the PS3 had. It's mostly "PC hardware", and PC hardware development didn't halt like the Cell did.
AMD's current GFX9 is ISA compatible with GFX7 GPUs, Zen is ISA compatible with Jaguar. Only thing that was dropped AFAIK are the Tensilica DSPs for audio, but it can probably be emulated by a more powerful CPU, or AMD can just throw the DSPs back in there for something like half a square millimeter at 7nm.
GDDR5 provides sufficient bandwidth with rather high latency, so there aren't the same problems as if it was using e.g. a large pool of eDRAM (but even then, Microsoft was able to circumvent that with GDDR5).
So unless Sony is going with a super wild (and super unlikely) hardware solution like an in-house GPU focused on RayTracing with A.I. co-processors and Power10 cores, the PS5 having BC is pretty much a given IMO.
Even more with the surge of Plus subscribers, everyone has a huge catalog of pretty good PS4 games to play, and that should be able to keep PS5 early adopters occupied during its (usually) slow first year.
Had the PS3 CPU been used for exclusively for non-graphics tasks, then it might have been easier to emulate, but the devs ran pixel shaders on the Cell due to a GPU that was lackluster in pixel shading performance at the time.
So Sony could either go with embedding a 3.2GHz Cell co-processor (whose development had ended in 2009) into their SoC that would either make it a lot more expensive to make or take away precious GCN GPU resources, or they could have thrown more money at further evolving Cell (a path they opted out very early according to Mark Cerny) or they could drop BC and have the higher performing and cheapest console launching in 2013.
The PS4 doesn't have anything as exotic as the PS3 had. It's mostly "PC hardware", and PC hardware development didn't halt like the Cell did.
AMD's current GFX9 is ISA compatible with GFX7 GPUs, Zen is ISA compatible with Jaguar. Only thing that was dropped AFAIK are the Tensilica DSPs for audio, but it can probably be emulated by a more powerful CPU, or AMD can just throw the DSPs back in there for something like half a square millimeter at 7nm.
GDDR5 provides sufficient bandwidth with rather high latency, so there aren't the same problems as if it was using e.g. a large pool of eDRAM (but even then, Microsoft was able to circumvent that with GDDR5).
So unless Sony is going with a super wild (and super unlikely) hardware solution like an in-house GPU focused on RayTracing with A.I. co-processors and Power10 cores, the PS5 having BC is pretty much a given IMO.
Even more with the surge of Plus subscribers, everyone has a huge catalog of pretty good PS4 games to play, and that should be able to keep PS5 early adopters occupied during its (usually) slow first year.
How small would Cell be at 7nm? I'd like to see it return just for the sake of Custom Hardware!
Using the power of The Internet:
Cell was 240M transistors and 220 mm^2 at 90nm. 7nm us 7 process nodes later, at half the size each node, would be 1/2^7 = 1/128th the original size, give or take. So 2mm.
C'mon, I'll happily donate 2mm of die space to an integrated CBE! Audio processing, ray tracing, physics, all in an efficient and totally unusable, dev-hated package.
Using the power of The Internet:
Cell was 240M transistors and 220 mm^2 at 90nm. 7nm us 7 process nodes later, at half the size each node, would be 1/2^7 = 1/128th the original size, give or take. So 2mm.
C'mon, I'll happily donate 2mm of die space to an integrated CBE! Audio processing, ray tracing, physics, all in an efficient and totally unusable, dev-hated package.
Last edited:
115 mm^2 at 45nm. 5 nodes. 70% size reduction each? 115 * 0.7^5 = 19 mm^2
Maybe a little over 5% of an SOC.
Maybe a little over 5% of an SOC.
What are they doing for PS Now? Last they talked about it, they were using the 45nm version.
Maybe the expense of shrinking the cell and integrating it in a PS5 SoC could make their rack units both lower power and universal compatibility with all generations of games. And consumer unit would naturally get it.
Maybe the expense of shrinking the cell and integrating it in a PS5 SoC could make their rack units both lower power and universal compatibility with all generations of games. And consumer unit would naturally get it.
Jason Schreier from Kotaku has released an article that sheds more light on current status of PS5. In short, it ain't coming anytime soon and their sources are generally uncertain about the shape and timing of next generation of hardware.
https://kotaku.com/sources-the-playstation-5-is-still-a-ways-off-1825152206
https://kotaku.com/sources-the-playstation-5-is-still-a-ways-off-1825152206
- Status
