• Hey everyone, staff have documented a list of banned content and subject matter that we feel are not consistent with site values, and don't make sense to host discussion of on Famiboards. This list (and the relevant reasoning per item) is viewable here.
  • Furukawa Speaks! We discuss the announcement of the Nintendo Switch Successor and our June Direct Predictions on the new episode of the Famiboards Discussion Club! Check it out here!

StarTopic Future Nintendo Hardware & Technology Speculation & Discussion |ST| (New Staff Post, Please read)

Reader mode
japtor said:
Or it was just an unsubstantiated rumor to begin with.

I feel like (for better or worse) caching/installs kinda have to be the way to go for faster performance. UHS-II is expensive, III basically doesn't exist on the market it seems (might as well be UFS), SD Express is a ways off (and will probably be expensive as well), and I don't see the game cards themselves becoming majorly faster either. Course with heavy usage like caching involves, you'd want big storage or expandable/replaceable memory to reduce the wear.

So I guess we're back to M.2 2230 seeming like the best option...size, power, and ruggedness concerns be damned.
Click to expand...
Given how throttled internal storage and the sd card are, there's room for growth. Given devs wanted 1GB/s storage in the other systems, I don't think the ceiling for Dane is as high as we thought
 
0
Dakhil said:
videocardz.com

Qualcomm Snapdragon G3x powers Razer's handheld gaming console, Snapdragon 8 Gen1 flagship SoC also leaked - VideoCardz.com

Qualcomm and Razer develop a Developer kit for next-gen mobile gaming The device, which is a Developer Kit, is powered by Snapdragon G3x silicon. We received a few slides featuring Qualcomm’s upcoming Snapdragon SoCs for 2022 products, including the G3x which was designed specifically for a...
videocardz.com videocardz.com

There's a rumour from FrontTron about the Snapdragon 8 Gen 1 being fabricated using Samsung's 4LPX process node, which is basically Samsung's 5LPP process node. Hopefully, Anandtech will have more details on which process node Qualcomm's using to fabricate the Snapdragon 8 Gen 1 later today.
Click to expand...
Samsung seems to be in a strong position with multiple rumours concerning their customers (AMD/Qualcomm for 3GAE, Exynos on 4LPE, 8GEN1 on 4LPX).
 
0


Hopefully Qualcomm will reveal the Snapdragon 8cx Gen 3 today (or tomorrow if not today), assuming that the Snapdragon 8cx Gen 3 uses the octa-core configuration of the Cortex-A78C based on the description of the CPU cores from a rumour.

And I'm curious about what CPU is being used for the Snapdragon 7c+ Gen 3.
 
0
ILikeFeet said:
Maybe Mochizuki will wake up today and choose violence
Click to expand...
So you are saying...
maxresdefault.jpg
 
MP! said:
edit wrong thread

But yes It's slow news ... I bet we won't hear anything until after the holidays and into the new year
Click to expand...
Eh, I do say there is a chance of us hearing something around TGAs next week.

But after that, yeah next year would be the next major infodump
 
Alovon11 said:
Eh, I do say there is a chance of us hearing something around TGAs next week.

But after that, yeah next year would be the next major infodump
Click to expand...
That's probably only in terms of software.

I think MP!'s also talking about hardware. And I personally don't expect to heard anything about hardware that's explicitly related to the DLSS model* until after 2021 ends.
 
Dakhil said:
That's probably only in terms of software.

I think MP!'s also talking about hardware. And I personally don't expect to heard anything about hardware that's explicitly related to the DLSS model* until after 2021 ends.
Click to expand...
software would still tell us something about the hardware at least
 
ILikeFeet said:
software would still tell us something about the hardware at least
Click to expand...
If Nintendo plans to release another trailer for the sequel to Breath of the Wild at The Game Awards 2021, the only obvious hardware clue that the trailer would give without engaging in a deep analysis is potentially Nintendo's planned launch window for the DLSS model* as of The Game Awards 2021.

And I'm talking about a notification about Dane being taped out and spec details when I said "...I personally don't expect to heard anything about hardware that's explicitly related to the DLSS model* until after 2021 ends."
 
Dakhil said:
That's probably only in terms of software.

I think MP!'s also talking about hardware. And I personally don't expect to heard anything about hardware that's explicitly related to the DLSS model* until after 2021 ends.
Click to expand...
yeah exactly more specifically I don't think, or expect rather, that someone like kopite would have any more Dane specific info until next year
 
0
Dakhil said:
I think as far as released commercial mobile Arm based SoCs are concerned, Samsung's 14LPE process node is the most advanced process node used for the Cortex-A57, which is one of the CPU cores that the Exynos 7420 uses. But I think TSMC's 10FF process node is the most advanced process node used for the Cortex-A57, which is used for TSMC's 10FF validation SoC.

And there was a rumour about Nintendo requesting a Tegra X1 sample from Nvidia, where TSMC's 7 nm** process node is used for fabrication, and the frequency of the Cortex-A57 cores increased from 1.9 GHz to 2.52 GHz, despite Nvidia and TSMC warning Nintendo about the difficulty of exceeding the Cortex-A57's max frequency of 1.9 GHz. And the power consumption was apparently higher than expected.

** → a marketing nomenclature used by all foundry companies
Click to expand...
The rumor seems believable to a degree, because as ILikeFeet mentioned, it would have offered the easiest way to deal with BC without having to do an extra turn.


And that is possibly disappointing if it can’t work for the 10/8nm family stack SEC. Could have been a nice little OS core.
 
ReddDreadtheLead said:
The rumor seems believable to a degree, because as ILikeFeet mentioned, it would have offered the easiest way to deal with BC without having to do an extra turn.


And that is possibly disappointing if it can’t work for the 10/8nm family stack SEC. Could have been a nice little OS core.
Click to expand...
I hope they aren't still considering a57s for the new device
 
ILikeFeet said:
It could be that 8gen1 is expected to be lower volume, allowing Qualcomm to make the design to produce as much yield as possible
Click to expand...
8gen1 already benefit from using a newer version of 5LPE (4LPX which is 5LPP) while E2200 is using 4LPE which is more in line with TSMC's 5 nm density (175 MTr/mm2).
 
0
Herb Alpert said:
They made this frankenstein soc for the WiiU for the sake of back compat, but I would hope they learned their lesson.
Click to expand...
Frankenstein chips and pushing existing designs to their limit are pretty common (see the PS5's design for the former, Wii and Gamecube for the latter). I don't think the Wii U's failure was about not getting the performance that Nintendo wanted - I think Nintendo got what they wanted, but what they wanted the market didn't.

Seeing if an uplock/dieshrink is viable seems minimum due diligence for an iterative successor. In a world where it panned out power wise, maybe we already have a switch pro, sitting at comfortable prices. I'm glad they didn't because I like seeing what else they come up with, but there would be some advantages...
 
Quoted by: SiG
1
Herb Alpert said:
They made this frankenstein soc for the WiiU for the sake of back compat, but I would hope they learned their lesson.
Click to expand...
I would expect a pandemic and a global chip shortage have induced significant delays that better alternatives have become available thanks to the improved processing nodes. I am also curious if the newer ARM cores and Nvidia's SoCs are capable of "hardware emulation" to a certain degree.

oldpuck said:
Seeing if an uplock/dieshrink is viable seems minimum due diligence for an iterative successor. In a world where it panned out power wise, maybe we already have a switch pro, sitting at comfortable prices. I'm glad they didn't because I like seeing what else they come up with, but there would be some advantages...
Click to expand...
I'd also say that Mariko does have the headroom for upclocks and performance boosts, but at this current situation I don't think the Mariko chips can be produced sustainably without incurring losses due to the whole TSMC pulling out discounts.

A lot of the decisions seems to be logistics, too. But like you, I'm in favor of seeing it so it forces Nintendo to upgrade to a more contemporary architecture.
 
I'd also prefer Nintendo and NV to tackle BC issues gradually over generations instead of relying on low-level hardware BC until it can't hold itself anymore and we have a Wii U redux. That said Wii U was a failure on so many fronts, that the one Nintendo took a lesson from is not necessarily the same as the one we would expect them to.
 
Quoted by: SiG
1
Mercury_Sagit said:
I'd also prefer Nintendo and NV to tackle BC issues gradually over generations instead of relying on low-level hardware BC until it can't hold itself anymore and we have a Wii U redux. That said Wii U was a failure on so many fronts, that the one Nintendo took a lesson from is not necessarily the same as the one we would expect them to.
Click to expand...
They should entrust backcompat hardware/software emulation to NERD, given their track record is pretty much up there with M2's.
 
SiG said:
I'd also say that Mariko does have the headroom for upclocks and performance boosts, but at this current situation I don't think the Mariko chips can be produced sustainably without incurring losses due to the whole TSMC pulling out discounts.
Click to expand...
I would really expect Nintendo to release a 8 nm version of TX1 with no cooling and a smaller battery in a smaller lite model for less than $100. Maybe in 2023 or later.
 
japtor said:
I feel like (for better or worse) caching/installs kinda have to be the way to go for faster performance. UHS-II is expensive, III basically doesn't exist on the market it seems (might as well be UFS), SD Express is a ways off (and will probably be expensive as well), and I don't see the game cards themselves becoming majorly faster either. Course with heavy usage like caching involves, you'd want big storage or expandable/replaceable memory to reduce the wear.

So I guess we're back to M.2 2230 seeming like the best option...size, power, and ruggedness concerns be damned.
Click to expand...
eUFS is still very much an applicable option for replacing eMMC for internal storage at least, given how widespread its usage is in Android phones, which would definitely be the option if internal installs are needed.

For comparison, the original Switch used a Toshiba THGBMHG8C2LBAIL eMMC chip, which was rated to the JEDEC eMMC 5.1 standard. The random read rating of eMMC 5.1 is up to 11,000 IOPS (input/output operations per second). Samsung‘s eUFS v3.1 chips (which are now in mass production) are rated at up to 100,000 IOPS in the same random read category. That’s nearly a literal order of magnitude increase in read performance. Write performance is a similar deal. Even eUFS v2.0 (from back in 2016) is rated at 4 times the random read performance of the fastest eMMC.

eUFS seems like the choice for more performant internal storage, the only thing that remains is what external storage will be. As we discussed prior, Nintendo could leverage its position to push for UFS card production.

As for game cards, we know so little about the game card and its reader’s theoretical upper limits, or if there are any other viable ROM-based options Nintendo could opt for other than Macronix‘s SIMD XtraROM solution.

Internal storage is all sewn up, as far as I’m concerned, it will be some version of eUFS. How Nintendo resolves the rest, if they choose to resolve it at all, is the only question that remains.
 
0
BlackTangMaster said:
I would really expect Nintendo to release a 8 nm version of TX1 with no cooling and a smaller battery in a smaller lite model for less than $100. Maybe in 2023 or later.
Click to expand...
the cost to reengineer the TX1 for 8nm would guarantee this costs $200+

davec00ke said:
Both the PS4 Pro and XBOX ONE SERIES X still had to have AMD Jaguar CPU's to keep compatibility
Click to expand...
and yet the successors have Ryzen 3000 style cpus and still have compatibility
 
davec00ke said:
They used Amd Jaguar cpu and a frankenstein GPU because it gave them easy compatibility
Click to expand...
Correct. Easy compatibility while meeting their requirements.

That's different than saying that a Pro would require it. Moving to a new architecture will cost them a lot more to achieve BC, but is doable.
 
0
davec00ke said:
They used Amd Jaguar cpu and a frankenstein GPU because it gave them easy compatibility
Click to expand...
the difference is that they had room to scale up. scaling is the path of least resistance. the CPU and GPU were already on the same nodes as the One and PS4, they just added GPU cores to them. Nintendo can't do that as they are space limited and the CPU and GPU were not made for such high clock speeds. moving them to 8nm would cost as much as R&D'ing Dane. maybe more as the costs aren't amortized like Dane is with Orin.
 
BlackTangMaster said:
I would really expect Nintendo to release a 8 nm version of TX1 with no cooling and a smaller battery in a smaller lite model for less than $100. Maybe in 2023 or later.
Click to expand...
That would require Nintendo and Nvidia to redesign the Tegra X1 with Samsung's IP in mind since the Tegra X1 was designed with TSMC's IPs from the beginning. And that alone would be one of the reasons for a console that costs more than $100 as ILikeFeet said.

But in the scenario that Nintendo and Nvidia are designing a 8 nm** variant of the Tegra X1, I hope Nintendo and Nvidia decide to at least use a noticeably better CPU (e.g. Cortex-A72).

davec00ke said:
Both the PS4 Pro and XBOX ONE SERIES X still had to have AMD Jaguar CPU's to keep compatibility
Click to expand...
Backwards compatibility shouldn't be a concern when the CPU is concerned, considering that the Armv8.2 architecture (assuming the DLSS model* uses the Cortex-A78C for the CPU, which runs on the Armv8.2 architecture) is backwards compatible with the Armv8.0 architecture (which the Cortex-A57, the CPU on the Tegra X1, runs on).

The problem when backwards compatibility is concerned is with the GPU.

~

Anyway, Qualcomm announced the Snapdragon 8 Gen 1 today. And Qualcomm only mentioned Samsung's 4 nm** process node was used. Qualcomm made no mention which variant of Samsung's 4 nm** process node being used to fabricate the Snapdragon 8 Gen 1. (There's a rumour from FrontTron about the Snapdragon 8 Gen 1 being fabricated using Samsung's 4LPX process node, which is basically Samsung's 5LPP process node.)
 
Speaking of PS4 Pro/Xbox One X sticking with Jaguar, the only realistic alternative in that time period would've been Puma, right?
Checking with wiki, I get the impression that Puma's improvements over Jaguar were suited for mobile, but not as meaningful for the Pro/X.
 
0
Dekuman said:
I hope they aren't still considering a57s for the new device
Click to expand...
Just to be clear, my comment was about using the A57 as the little cores or OS cores. They offer a similar level of perf to the A510 at the same clock (theoretically).

davec00ke said:
Both the PS4 Pro and XBOX ONE SERIES X still had to have AMD Jaguar CPU's to keep compatibility
Click to expand...
They didn’t need the Jaguar CPUs to have compatibility.

and I’m not sure what console you are referring to in the bolded
davec00ke said:
Its a solvable problem for a Switch 2 but probably not worth the hassle for a Switch PRO
Click to expand...

It was going to be a hassle regardless if it was the pro or the 2.
 
ReddDreadtheLead said:
and I’m not sure what console you are referring to in the bolded
Click to expand...
Probably only the Xbox One X since davec00ke mentioned a Jaguar CPU. (Microsoft isn't exactly doing any favours by using 'X' for two different consoles, with the only difference, name wise, being 'One' vs 'Series'. The same could be said about Microsoft using 'S' for two different consoles.)
 
Dakhil said:
Probably only the Xbox One X since davec00ke mentioned a Jaguar CPU. (Microsoft isn't exactly doing any favours by using 'X' for two different consoles, with the only difference, name wise, being 'One' vs 'Series'. The same could be said about Microsoft using 'S' for two different consoles.)
Click to expand...
Sorry I mean Xbox One X
 
0
Reader mode
You must log in or register to reply here.
Please read this new, consolidated staff post before posting.

Furthermore, according to this follow-up post, all off-topic chat will be moderated.
Last edited by a moderator:


Back
Top Bottom