Next Generation Hardware Speculation with a Technical Spin [2018]

Status
Not open for further replies.
SSDs should also be a lot more durable, so they'd probably have to do somewhat less repairs.
Let’s hope we get a drive with good endurance.

Still, they’re pushing the layer count up as well as going to QLC. Those big leaps in storage per unit cost just aren’t available in magnetic media anymore.
 
But will 1TB be enough for 2021 and later?

Has anyone said if all assets in the install files are unique or if there is often reasonable levels of duplication to allow for contiguous reading for performance over storage effecieny?

Ssd may not need such optimisation (if it exists)
 
But will 1TB be enough for 2021 and later?

I think so but I think we should start with 2TB.

This generation marketed a big increase in sizes because of PBR and baked lighting.
If we get rid as much as possible of baked lighting we may not see a big increase.

Textures and models are already pretty good in most games.

What we need is much more RAM to old more high quality assets at the same times and 4K resolution as standard.

I think at least some sort of SSD support is mandatory, maybe to to be used to store the last 1-2 games at least
 
Samsung just announced their new QVO SSDs. 1TB for $149, 2TB for $299, and 4TB for $599. It wouldn’t surprise me if by 2020, you could get a TB for half that price at retail.

I think the chances are increasing that next gen will come with a SSD as the costs are getting reasonable and will scale down better than hard dives.

That’s one of my next gen wishes: a default 1TB SSD drive.

I think the SSDs with QLC are still very overpriced because they're just entering the market and NAND makers haven't made any price distinction between them and their TLC+SLC brethren, much less the models with TLC+DRAM.
For example, on Amazon UK we can find the 2TB Crucial MX500 for £210 (~$270) right now and I doubt Amazon or Crucial are losing money when selling at that price.
Price-per-GB right now is pretty much the same between the models with TLC+DRAM (Crucial MX500, 860 Evo), TLC+SLC (Crucial BX500) and QLC (860 QVO). According to projected production costs, they should all be selling for different prices (the first more expensive than the second, second more expensive than the third).
Maybe it's because SSD makers are trying to capitalize on the new "brands", or because it would be worthless to try to differentiate during the black friday / cyber monday season, or QLC yields aren't great at the moment.


That said, I doubt very much that SATA QLC drives will cost 15c/GB even on the short term. IMO they're likely to go towards <10c/GB next year and even lower in 2020. And that's final price for a consumer product, of course.
I wouldn't be surprised to see a $500 next-gen console launching with a M.2 128GB NVMe using PCIe 4.0 to serve as large cache for a M.2 2TB QLC SATA drive (no need to spend space on normal 2.5" drives... the PCB size of the 860 QVO is ridiculous).
 
Games just have gotten very big, this with 50GB patches and DLC, i think 2TB atleast would be nice, so that people dont have to swap drives or add external drives on a large scale.
 
As Shifty said before, you need to be able to guarantee the developers a sustained performance level for next-gen game experiences, which is not possible if you also want to allow for people to bring their own external drives to the system unless the SSD is a cache style device. Using SSD as only a dumb bulk install drive is a mistake. We don't need that quantity of fast storage pools.
 
Intel's 660p SSD is cheap [4bits per cell QLC], it's large, and it has 5 year warranty with 100-200GB per day usage [depending on the drive size], which is MORE THAN ENOUGH for vast majority of console player [game installations do not happen often].

If SSD land in consoles, IMO it will have QLC memory.
 
I'm mildly optimistic that a 128-256GB nvme drive will be in next gen consoles, as well as a storage solution, whether that be flash or magnetic.

That's my preferred scenario. I think it's most likely too, because it gives Sony and MS flexibility to launch with, say, a 128GB NVME and 1 or 2TB HDD, and make the change to SSD when the same capacity's cheaper.

On a different topic, I was reading the "designing a perfect controller" thread and I wonder if anyone here agrees or disagrees with my preferred DualShock 5:

1) Imagine the Joy Cons crossed with the DualShock 4.

2) Each half can emit light somehow e.g. IR LED's.

3) They can be connected via a central unit to resemble the DualShock 4.

4) Left controller, right controller, and central unit comes with each console.

5) This central unit contains one camera, microphone, and speaker; several cameras, microphones, and speakers; or some combination thereof.

6) When the left, right, and central sections are combined, the camera or cameras are used for tracking said combined device's position.

7) When detached, the central section can be physically connected to the console via USB-C and the camera or cameras used to track some quantity of left and right controllers.

8) Multiple central sections can be connected to a single console in order to improve tracking.

I'd also like the central section to have a simple, cheap, low resolution touchscreen. Maybe black and white OLED. Enough to present some extra buttons, shortcuts to items, or a HUD.

It'd render every console capable of full motion controls out of the box, and make entry into VR even cheaper. If the PS5 itself does away with the need for a breakout box, this controller would do away with the need to buy a camera and motion controllers.

The corollary of that is the expense of such a controller, relative to that of a tarted up DualShock 4, is passed on to every customer. But it seems to have worked quite well for Nintendo and their Joy Cons.

There's also the benefit that plenty of developers could get two or three player games out of a single console's default controller. Especially with the Switch being on the market.
 
As I understand it, there's an expectation that prices are going to drop faster than they have because price fixing has been addressed. A launch console with a $50 SSD versus a $30 HDD is also viable as a loss-leader as that same flash will drop to $25 and then $15 within a couple of years, possibly. It's also smaller and lighter than a drive, allowing for a smaller console with additional savings through the retail chain per unit.

Remember too if you go without a disc drive you will have even more money to play with. A 4k bluray drive is what $20-$40. If you do that off the bat you can fit a pretty big ssd in the console. I bought a 960 evo for $120 on black Friday 1TB and I got a wd black 1tb (nvme) for $180. A 2020 console should be able to start with one of them. For those who want a disc drive they can get a disc drive with less storage for more money.
 
If you do that off the bat you can fit a pretty big ssd in the console. I bought a 960 evo for $120 on black Friday 1TB and I got a wd black 1tb (nvme) for $180. A 2020 console should be able to start with one of them. For those who want a disc drive they can get a disc drive with less storage for more money.

If you look at IHS (iSupply) breakdowns for consoles you'll see that the HDD has never cost more than 10% of the RRP and it's usually much lower. I don't see decent sized SSDs dropping that fast even by 2020. If games continue to trend upwards of 100Gb, 1Tb is going to be your realistic minimum with 4Tb (or more) for keen gamers.
 
I don't know that cost at release is as important as cost over the life of the generation.

I'm pretty sure the current HDD cost is more than 10% of the current cost of a console
 
Might be some value in discussing SSD.
https://gearnuke.com/microsoft-looking-staff-leads-next-gen-xbox-development/
We are seeking a qualified candidate for a Senior Engineer to develop storage solutions with the Xbox console hardware development team. Storage solutions include HDD, SSD, and Flash based storage technologies. The successful candidate will be the primary technical focal point for existing and leading-edge storage technologies on currently shipping and future Xbox design projects.

Sounds like they're getting further along here with Xbox Development, they would have been totally blind-sided by a Sony 2019 release. I guess they must have known it wasn't going to happen.
 
If intel wins one of the console deals I expect there to be significant amount of optane as cache. It could also help with ram as some stuff might always stay in cache instead of ram.
 
Or they don't care. It's insane business to try and direct your products and services based on trying to outmanoeuvre your opponent. MS have their plan and they'll stick to it. Sony have theirs and they'll stick to it. They then see what the other does and work out competition based on what players they've fielded.
Wouldn't say don't care, but it's a lot lower on list of priorities.
Design the best device, at a given price point and release time frame and take it from there.
Everything else is marketing and software/games.

They both have a reasonable idea what the other can do at specific price points and time lines.
One of the reasons I hope one of them does something architecturally different
 
If intel wins one of the console deals I expect there to be significant amount of optane as cache. It could also help with ram as some stuff might always stay in cache instead of ram.

I think 3D Xpoint is intriguing as well for this application, but I think Micron is a more likely supplier.
 
Status
Not open for further replies.
Back
Top