LOL, thank you.  I needed a good laugh.

I meant things like the drive configurations and so on that everyone is commenting on without us actually having all the facts.

There is still allot of unknowns at this point.... Like how powerful the console is going to be which everyone seems to be throwing ideas around on. (Like dedicated Ray Tracing hardware and 80 CU's.)

I am not saying your assertion is either correct or incorrect, merely just asking for evidence.

Xenogears? I feel like you should worry more about whether PS1 is even included in the PS5 BC first

Even if they did it wouldnt be what we had on ps3/ psp. There's a very good chance the games were only licensed for those emulators and any modification nulls the licensing.

I imagine when they began the ps4 gen that the idea they would move forward with bc onto ps5 was not something being discussed. But since it is happening i csn see them doing another emulator (for ps4 with ps5?). But the games are going to have to be licensed again so it will be a while till its tbe same library.

Well I think everyone here is making an assumption based on what has been provided by Cerny and how it could be achievable, given the available and upcoming tech for PC's. Sure Navi is based on GCN and it will be probably Navi 10 also AMD's focus with Navi was to fix GCN previous shortcoming and make it more efficient. Yet We don't know anything what Navi has to offer and many reports suggest Navi will have an answer Nvidia RTX, whether Raytracing support is achieved by asynchronous compute or has an dedicated Hardware remains to be seen 7/7

Totally. That is the ideal situation but it seems unlikely to go that far. PS4 discs might even be a bit of a question mark.

BC is great news, but definitely we need to see how it is all done.

Graphics Core Next is efficient. It's just not efficient at gaming, it's a very compute centric design... And because of such only shows it's true potential when backed with inordinate amounts of memory bandwidth to keep everything fed... Which is where Asynchronous compute came into play.

Navi is likely to be a Polaris replacement... A Mid-Range part, but implement some of the newer tricks found in Vega like Draw Stream Binning Rasterization, Primitive Shaders and so on. (And probably make them work the majority of the time!). - It will also likely see the bulk of it's performance gains with a re-balancing of the functional units to achieve more throughput in gaming orientated workloads and a couple of new tricks. But it's still Graphics Core Next with all it's Pro's and many many many Con's.

I doubt AMD will take the approach nVidia has with Ray Tracing and include components specifically for that task (RT Cores) that cost transistors, rather they will do what they have historically done and rework their shader pipelines to make them more effective with those tasks...
Remember we have been down a similar path before... nVidia with the Geforce FX/5 spent a ton of transistors on retaining fixed-function hardware like TnL, whilst AMD with the Radeon 9700 series spent it's transistor budget on more general-purpose hardware... And well. The rest is history.

Vega 7 for instance is able to do 1 Gigarays compared to the Geforce RTX 2080 8.

And in Claybook Vega 64 is doing 9.76 Gigarays which is a pretty impressive feat.

So I think AMD will leverage GCN's inherent compute strengths and simply make it more flexible to lend itself better to Ray Tracing. But that's an assertion at this point, I don't think AMD has the turn-around time to bake dedicated Ray Tracing hardware into it's GPU's on such a short notice, maybe with it's Next-Gen hardware.

I think we should add a few frontloaded expenses that change your rosy view:

a) Cost of making prototypes. Given how extremely expensive a large 7nm design is (currently), and assuming Sony actually has some working prototypes of PS5s (i.e. they are not high-specced PCs, we can very roughly add >=500M for developing the SoC (and whatever goes with it, peanuts in comparison to the SoC).

b) Every mask failure adds xyM costs to the development cycle.

c) Making, say, functional developer units at 25k a piece adds another xyzM to the bill

d) Final masks for the PS5, producing an initial run of consoles = 100M

e) Actually paying for PS+ (Sony gets a fraction of what you pay with your PS+ subscription, most is for renting hardware/maintaining own hardware and software people)

f) Adjust for money you don't actually can invest if you sell consoles at a great loss instead of break-even on it.

g) All the stuff I forgot to add to the list (like demo units for shops, etc, new surround heaphones, PSVR2, etc. etc.). The list would probably go to z)

Dude, that's an insanely disingenuous comparison. You're comparing the fringe discounts of a 3 year old model to the one-day-old announcement price of a new digital-only model without discounts. If they were launched at the same time, the implication to be made is that it still would have been $50 less, just like it is now for the undiscounted price. Had this model been 3 years old like the current XBO S, it would likely be just as easy to find at $170. I don't see how PS5 would be any different.

https://www.resetera.com/threads/official-playstation-next-gen-plans-2020-wired-exclusive-article-ray-tracing-bc-ssd-more.111777/page-73#post-19929295

Post #3648

By now, it should be obvious that I don't post rumors out of nowhere. ^_^

8 GB of RAM for the OS would be a bloated OS. No thanks. Switch's OS is the best OS, because it boots up instantly, just plays games, and only takes up 300 KB of space. 

I meant things like the drive configurations and so on that everyone is commenting on without us actually having all the facts.

There is still allot of unknowns at this point.... Like how powerful the console is going to be which everyone seems to be throwing ideas around on. (Like dedicated Ray Tracing hardware and 80 CU's.)

I am not saying your assertion is either correct or incorrect, merely just asking for evidence.

To be fair, my original assumption was for a device that targeted a $400 USD price point.
Consoles that are priced higher like the Xbox One X obviously can spend more on various components.

Ray Tracing has been possible for years, it really depends how extensive you wish to implement the technology, the more robust your implementation of Ray Tracing is... The more hardware you need to throw at the problem.
Digital Foundry also recognizes this, hence their claim why it's technically possible to do on a base Xbox One.

And to be fair... Games were starting to dabble in Ray Tracing even on the 7th gen with a few path traced games in a few deferred renderer engines.

Indeed. It is certainly a good thing, competitive pressure from Microsoft has likely assisted to that end... As original Sony was fairly dismissive of the idea of backwards compatibility for one reason or another on the Playstation 4.

https://www.extremetech.com/gaming/250579-everybody-always-claims-want-console-backwards-compatibility-doesnt-anybody-use

I would like a full extensive roll out though across every Playstation console, not a curated approach like Microsoft. - Both approaches have their Pro's and Con's obviously, but I digress.

I concur. GDDR6 seems to be the more economical choice right now. But hey, if it has HBM2... I will be pleasantly surprised.

We have absolutely no idea about the drive configurations. - For all we know it's a chunk of NAND soldered onto the motherboard with a PCI-E 4.0 interconnect that caches a spinning mechanical hard drive... And such an approach would possibly not just cache an internal mechanical hard drive.. But potentially external mechanical drives as well.

We aren't at a point where we can install all our games on NAND... Heck, have 13 Terabytes total on my Xbox One X and I am thinking about boosting even that amount.

Real world performance will be more. The Xbox One X chip isn't using the latest and greatest of Graphics Core Next... So there is efficiency gains to be had with Navi, heck even Vega.

By 2020 7nm should be far more mature than it is currently, so Sony and Microsoft may be willing to invest in larger chips initially.

Not as simple as that I am afraid. The chips themselves have a correlated increase in cost with increases in size.
The Xbox One's APU was roughly the same size as the Playstation 4's yet substantially inferior in terms of overall capability. - That is because of the ESRAM took up a chunk of the chip size.

The thing with using NAND as a high-traffic swap cache is the idea of one write, many reads.
Which is why Readyboost cache was a thing, which is why the Corsair Accelerator existed and why the Sandisk Radycache was rather popular for awhile.
Add onto the mechanical drives with a chunk of NAND like the Seagate Firecuda SSHD.... And it's actually a common idea.

Optane is NAND which is optimized for it's task, rather than using commodity chips.

Wut

You re welcome. I can laugh about the staggering ignorance people like you chose to display.