• 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.

StarTopic Future Nintendo Hardware & Technology Speculation & Discussion |ST| (Read the staff posts before commenting!)

Ugh, Spawn Wave reporting on Switch NG. Why do these clickbait news sources/YouTubers report on fake leaks after they've been publicly debunked? Like I get rushing out to report on them before they're confirmed but why do this?
What else but views?
 
0
If everything in this analysis by Thraktor pans out

3.4 TFlops docked/1.7 TFlops handheld. Still plenty good, plus we have more RAM, have FDE, DLSS, better architecture.
If Thraktor's predictions are even remotely on the money I will cry ray traced tears of happiness in real life.

Even 1.5 tflops handheld and 3 tflops docked would make me happy.

2 tflops handheld and 4 tflops would require me to fly to Nintendo HQ and shake everyone's hands, then proceed to get arrested.
 
I don’t think cutting a few ounces of battery weight would be a very appealing product. The Lite for the Switch 1 already sold not great so making a Lite that has a much worse battery and is only a couple bucks cheaper and a couple ounces smaller could really struggle.

Well, my thinking is Switch 2 weighing ~500 grams (with the ~8" screen and bigger joycons, with decent ergonomics this time) at $400.

Switch 2 Lite weighting ~300g (with a 6"-6.2" screen (I think they can fit at least a 6" screen with something the same size of the current Lite) and the same flat design) at $250.

Even then, I still believe Switch 2 would sell much, much more than the Lite. People just loved the hybrid flexibility. But having more options is always good.
 
0
The Gamescom leak means we know there appears to be a very performant product in the works despite not knowing #SM or #nm.

Not that chaotic.
 
Thinking about, the timing doesn't make much sense at 8NM either, and I'm not talking about the device being dated for 2024.

If we are indeed talking about an 8NM chip closely related to Orin, then the SoC would have been tapped out and ready for manufacturing about two years now, if not three, and Nintendo has been sitting on it. Yes, that could happen, but then the rollout would have been much smoother. Nintendo would have ample time to decide on final power profile, as well as RAM setup. Dev kits would been plentiful and ready to go for quite some time. Important studios like Activision would have known the final specs probably by the end of 2022, if not earlier. The Matrix demo would have been running on finalized hardware rather than "target specs".

Also, Nintendo would had ample time for logistics and preparing software for the device. Therefore, It believe there would a much less uncertainty about release date.
 
Last edited:
introducing the switch 2 official carrying case


ahjxe87lsin91.jpg


s-l1600.jpg
 
Thinking about, the timing doesn't make much sense at 8NM either, and I'm not talking about the device being dated for 2024.

If we are indeed talking about an 8NM chip closely related to Orin, then the SoC would have been tapped out and ready for manufacturing about two years now, if not three, and Nintendo has been sitting on it. Yes, that could happen, but then the rollout would have been much smoother. Nintendo would have ample time to decide on final power profile, as well as RAM setup. Dev kits would been plentiful and ready to go. Important studios like Activision would have known the final specs probably by the end of 2022, if not earlier. The Matrix demo would have been running on finalized hardware rather than "target specs".
T239 was taped out and ready for production within the same timeframe as Nvidia's other 4N products, so that's another notch in 4N's belt.
 
Thinking about, the timing doesn't make much sense at 8NM either, and I'm not talking about the device being dated for 2024.

If we are indeed talking about an 8NM chip closely related to Orin, then the SoC would have been tapped out and ready for manufacturing about two years now, if not three, and Nintendo has been sitting on it. Yes, that could happen, but then the rollout would have been much smoother. Nintendo would have ample time to decide on final power profile, as well as RAM setup. Dev kits would been plentiful and ready to go for quite some time. Important studios like Activision would have known the final specs probably by the end of 2022, if not earlier. The Matrix demo would have been running on finalized hardware rather than "target specs".
Node doesn't affect timeliness like that. They can fab an 8nm design whenever as long as Samsung still has equipment
 
Is there a path from Samsung 8nm for a future revision of the Switch 2 like there was with the Switch going from the TX1 to TX1+ (20nm to 16nm)? If not, then it really wouldn't make sense to go with that, right? We're talking about an SoC that is going to be in use for the next 6+ years. Nintendo didn't have any real control when they first contracted the use of the Tegra X1, but with the T239 being custom with budgets and expectations on their behalf, they have a LOT more.
Hypothetically yes. But that requires redesigning the SoC for any process node newer than Samsung's 8N process node, similar to how Nintendo and Nvidia redesigned the SoC when migrating from TSMC's 20 nm* process node (Tegra X1) to TSMC's 16FF process node (Tegra X1+) by updating the RAM controller from a LPDDR4 controller to a LPDDR4X controller.
* → a marketing nomenclature used by all foundry companies
Is there a road from 4N to N4 or 3nm process?
Hypothetically yes, for transitioning from TSMC's 4N process node to TSMC's N4 process node, since both are part of TSMC's N5 process node family, and therefore share the same IP. But considering TSMC's 4N process node's specifically customised for Nvidia, and TSMC's 4N process node is very similar to TSMC's N4 process node, I don't really see the point.

And hypothetically no, for transitioning from TSMC's 4N process node to TSMC's N3E process node, without redesigning the SoC, since TSMC's N3E process node is not IP compatible with TSMC's 4N process node.
 
y'know what

if it does turn out to be a steam deck sized or bigger monster as has been rumored I'm 100% making it an "it's so bad" briefcase

bookmark this post
 
What's one consider decent battery life? Switch OG?

Yes.

And that will be hard to get to at 22W unless you’re doing a monster 7000 mAh battery.

We’ll see the power draw and size this clocks in at, it’s not a huge deal though it being huge would add some cost to the end user with increased production costs.
 
0
Node doesn't affect timeliness like that. They can fab an 8nm design whenever as long as Samsung still has equipment
I'm saying that the device would have been taped out and production samples ready to go, not that mass production started 3 years ago. I doubt that Nintendo/Nvidia tapped the chip in 2020/21, and then did nothing with it for a couple of years.

Edit: I guess another possibility is that Nintendo didn't do any kind of RnD after Mariko, then rang Nvidia in late 2021, at which point Nvidia went "well, since you are in a hurry, I have this Orin we could repurpose".
 
Last edited:
What's one consider decent battery life? Switch OG?
Personally, 3 and a half hours of battery life for a handheld system is my expected minimum. The Steam Deck pushed my thoughts a smidge given it's a handheld PC with a lot more weirdness with it's architecture, but generally... yeah 3 and a half is my absolute minimum.

Above that is much better though. "Minimum" is not "I think this is ideal"
 
A local man died on Thursday while playing with the latest Nintendo console in bed. Daryl Roberts, 37, had just got the Nintendo Switch 2 and Breath of the Wild Deluxe when they launched last Friday and was looking forward to some quality time with the game. Unfortunately within what forensic specialists estimate to be no longer than fifteen minutes one or both of Roberts' hands fell asleep. Roberts then dropped the device which, due to its substantial size and weight, crushed his skull, killing him instantly. A representative from Nintendo reminded customers that while the new iteration of Nintendo Switch still enables home console gaming on the go, it is not meant to be lifted above the shoulders, held for longer than an hour, or be used by children under the age of twelve. Nintendo recommends players use the system's "tabletop mode" and to make sure the tabletop used is not made of particle board or any other composite.
 
Last edited:
Having a 1/2 split would be ideal, you don't want docked to be ridiculously ahead of handheld and viceversa. If it is really 4 TFLOPS when docked, handheld must be 2 or close to 2.
Why though? If the Switch 2 is expected to play 3rd party games that are on 10.2tf PS5s and 12tf XSXs, Switch 2 can and should go highest possible imo. It won't matter the 1/2 split. People will still see the Switch 2 portable / PS5 / XSX difference. Imo try and make docked as close to those other 2 home consoles as possible.
 
Is the primary debate about node based on concern/wonder of efficiency?
I think it's a bit of concern of performance and incredulity over the leaked SoC being fabbed at 8N, which i guess does go back to performance (e.g. the 12SM chip at 8N would be large inefficient at very low clocks at 8N which may suggest we were completely wrong and a different design with less cores is being used if it is indeed 8N)
 
Last edited:
0
Is the primary debate about node based on concern/wonder of efficiency?
I'm surprised so many people are responding with battery life. IMO most people are worried about it because they're worried about the peak clock speed, i.e. performance. The first Switch is on a pretty inefficient node, so Nintendo had to use low clocks to hit the desired heat and/or battery life. The V2 revision is on a more efficient node, but they didn't want to create a performance difference at that point, so they "spent" the node increase entirely on battery life. If they'd started on the better node in 2017, they could have balanced it between somewhat higher clocks while still having better battery life than the V1 we got. So that's how I think people are seeing the node debate now, a chance to start on the "good" node.
 
A local man died on Thursday while playing with the latest Nintendo console in bed. Daryl Roberts, 37, had just got the Nintendo Switch 2 and Breath of the Wild Deluxe when they launched last Friday and was looking forward to some quality time with the game. Unfortunately within what forensic specialists estimate to be no longer than fifteen minutes one or both of Roberts' hands fell fell asleep. Roberts then dropped the device which, due to its substantial size and weight, crushed his skull, killing him instantly. A representative from Nintendo reminded customers that while the new iteration of Nintendo Switch still enables home console gaming on the go, it is not meant to be lifted above the shoulders, held for longer than an hour, or be used by children under the age of twelve. Nintendo recommends players use the system's "tabletop mode" and to make sure the tabletop used is not made of particle board or any other composite.
do-you-even-lift-bro-pumping-iron.gif
 
I'm saying that the device would have been taped out and production samples ready to go, not that mass production started 3 years ago. I doubt that Nintendo/Nvidia tapped the chip in 2020/21, and then did nothing with it for a couple of years.

Edit: I guess another possibility is that Nintendo didn't do any kind of RnD after Mariko, then rang Nvidia in late 2021, at which point Nvidia went "well, since you are in a hurry, I have this Orin we could repurpose".
it depends on when they started the development. this wouldn't change, whether it's 8nm or 4nm. and since this SoC still has parts that aren't in Orin, we know that it's only very loosely based on Orin. it shares CPU and GPU arch, but that's about it. it still wouldn't make the design go any quicker
 
0
Can we ban posts about Nate's voice? Apparently tagging people is punishable by death but simping over (let's be real) a completely average voice for pages at a time doesn't involve moderation?
I actually do not mind it, I am pretty good at scrolling past posts that are nonrelevant to the conversation. I do not mind most things posted here but I totally get why people get bent out of shape or frustrated when people are randomly talking about the next Smash's roster or when the next Bayonetta will come out which has nothing to do with Nintendo's next-gen hardware. It just dilutes the thread.

What DOES drive me up the wall is the somewhat arbitrary basis on what is and isn't on topic, or when people strip any nuance from your post and tell you to carry that conversation in another thread. This can be done to posts regarding many of Nintendo's business practices or patterns which are discussed ad nauseam in here. You correctly pointed out that Nate's voice which is well-suited for his line of content creation has nothing to do with the NG (unless he's performed secret voice work for Nintendo which would result in his voice greeting us every time we turn on the NG).

But apparently this:

In an ideal world, Pokemon fully utilizing NG's hardware would be a step above what we're seeing with Genshin Impact. I think generally speaking, open world games will receive a big bump up on NG due to the CPU increase and the memory bandwidth bottleneck being alleviated.



Just imagine these environments sprawling with Pokemon and other trainers or landmarks. Pokemon swimming, flying, galloping, and what have you. The world would be more interactive as well. Not quite BoTW/ToTK but just enough to be engaging.

As for Elden Ring that doesn't even run on PS5 at a smooth 60fps so if it did come to Drake, I can see it being solid running on NG at 30fps but wouldn't be shocked to see a performance mode of some kind. I don't have Elden Ring on PC so I am not sure how well it scales and the most taxing parts on the HW.
Is me being off-topic and derailing. So I have to ask, why do we care about the silicon in NG if not for what we hope or speculate what it will be capable of? I know GameFreak ain't the most talented when it comes to the technical side of things but my whole point was that the baseline for open-world games on Nintendo platforms goes up by default, not that GameFreak is shit and that Pokemon looks like a PS1 game or whatever. People will say those things but I can't control that the same way we cannot control journalist sites creating fake news based on the speculation that goes on in here.

That's all. If that is derailing, then 90% of the thread can be seen as derailing IMO. I just ask for more consistency because I'd be totally down for a "NG Graphics- Visual Technologies & Capabilities" thread if it is needed.

I think backseat modding and course-correcting the discussion are totally valid when people start posting AI art or begin ranking 3D Mario games. Speculating what kind of software should be able to scale down to NG should be fair game. The Gamescom news telling us that this thing can run the Matrix demo well and BoTW at 4k/60 is the news that told us more about the hardware than a majority of the aimless heavily speculative discussion that takes place here from time to time. This is no shade at the people who do good work here and keep the chains moving in regard to worthwhile discussion and speculation. This is just what I've noticed in the past month or so.
 
personally I actually want to take it places without getting it its own foam-lined briefcase

y'know, like the one the "it's so bad" kid puts his power glove in
Do you really think Nintendo is going to put out a product that generates an unsafe amount of heat?
 
Okay so I’m talking about the relentless scalping issues in Japan that seem to be strongly affecting the PS5 is the thing. Software sales are just super strangely low relative to hardware sales and those sales plus the extremely weak yen seem to point to PS5s being shipped out of Japan.

This is a pretty serious issue for Nintendo if they want the Switch 2 to have a large Japanese user base for games like Dragon Quest etc.
The device was going to be scalped to high hell regardless of the price. What deters the scalping is availability not the price, hard to come by the more enticing it is to be scalped. Furukawa already made a comment about this subject in that shareholder Q&A meeting a few months ago

And the PS5 scalping was reduced when Sony started to produce more of the console to the market. It’s not about price, it’s about being available. Making it more expensive helps very little.
 
Why though? If the Switch 2 is expected to play 3rd party games that are on 10.2tf PS5s and 12tf XSXs, Switch 2 can and should go highest possible imo. It won't matter the 1/2 split. People will still see the Switch 2 portable / PS5 / XSX difference. Imo try and make docked as close to those other 2 home consoles as possible.
That's not how it works all around. No developer is going to target docked mode from the get go, everything has to run on the handheld mode no matter what. Having a massive split would mean disparity between modes, two different platforms to develop for... And that's something Nintendo sure doesn't want them to deal with.
 
Quoted by: TLZ
1
Is the primary debate about node based on concern/wonder of efficiency?
Pretty much a bit of both?


Like LiC said
I'm surprised so many people are responding with battery life. IMO most people are worried about it because they're worried about the peak clock speed, i.e. performance. The first Switch is on a pretty inefficient node, so Nintendo had to use low clocks to hit the desired heat and/or battery life. The V2 revision is on a more efficient node, but they didn't want to create a performance difference at that point, so they "spent" the node increase entirely on battery life. If they'd started on the better node in 2017, they could have balanced it between somewhat higher clocks while still having better battery life than the V1 we got. So that's how I think people are seeing the node debate now, a chance to start on the "good" node.

But I would also add that it has shifted a bit towards Battery life/Heat concerns as of late after the Gamescom Leak.

Primarily because of the very well-known inefficiencies of Samsung 8nm combined with the sheer size of the SoC.

They have a full Desktop Ampere GPC in there rather than a custom GPC layout. So, power consumption on Samsung 8N would be as such without dramatic improvement to the Architecture, and size would be a bit tight for a tablet form-factor, likely pushing past M1 territory in die-size.

Which considering M1 is already 119mm2...on TSMC 5nm, and it uses ARM's (at least iirc) bigger A78C cores, would likely push it beyond the die size of Lockhart (Series S's SoC).

And that is a best case just factoring them dividing GA107 by 2 to get 1GPC's die-size and 8 A78C without considering stuff outside of those two on the SoC like the File-Decompression Engine. Or that the die size of a single Ampere GPC on Samsung 8N would very much be bigger than 100mm2 due to parts of the die that wouldn't shrink evenly and especially SRAM scaling hitting hard (Ditto that the SoC would have more cache outside of the GPU for it's size due to the A78C and a SysLC being likely as-per Standard design of ARM SoCs)

So you'd have a power hungry hot and hard-to-cool SoC if they went with Samsung 8N unless NVIDIA pulled all the stops on power efficiency. Even if they clocked it as low as they could go.

Which the Gamescom leak suggests they in fact didn't go low clocked as it ran Matrix Awakens with all RT effects enabled at a high enough resolution and quality output to actually eye up similarly or better than Series S's iteration of the demo.
 
Last edited:
Please read this staff post before posting.

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


Back
Top Bottom