• 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!)

I really hope we get some new information coming out of the GDC, while the leak was fun it stills feel like there is so many other missing components
We need a new Bloomberg article to hit. Starting last March it felt like there was a new article on the next Switch every 2-3 weeks.

If it's coming this fall, Bloomberg's sources will start hearing about things going into motion on the manufacturing side.

If we get to May though and haven't heard anything then I doubt it's coming this year.
 
We need a new Bloomberg article to hit. Starting last March it felt like there was a new article on the next Switch every 2-3 weeks.

If it's coming this fall, Bloomberg's sources will start hearing about things going into motion on the manufacturing side.

If we get to May though and haven't heard anything then I doubt it's coming this year.
Bloomberg probably wants to double/ triple/ quadruple check before they release another article.
 
Bloomberg probably wants to double/ triple/ quadruple check before they release another article.

Can’t wait for the amount of snark they’ll throw in this one.

“At least 20 developers now working on titles for a 4K Switch that does not exist. Many aim to release these new titles on the non-existent platform by late 2022.”

More realistically they’ll probably start talking more concrete about production and release timing. Seems like that’s gotta be around the corner
 
Can’t wait for the amount of snark they’ll throw in this one.

“At least 20 developers now working on titles for a 4K Switch that does not exist. Many aim to release these new titles on the non-existent platform by late 2022.”
This will be followed by an official response the next day by Nintendo vaguely denying the claims. Then an hour later the US patent office will publish a patent submitted by Nintendo for an in-house version of ray tracing.
 
If a new model is coming this year then we should hear some rumblings of it at GDC right? If nothing comes out of that it’s likely nothing will be launching in 2022?
 
If a new model is coming this year then we should hear some rumblings of it at GDC right? If nothing comes out of that it’s likely nothing will be launching in 2022?
Ehh, I say moreso Nintendo's next finanical meeting covering FY23 would be the make or break for Drake 2022/2023
 
I've been out of the loop but are people expecting a new model to launch this year? I was assuming 2023 for some reason.
 
Thanks Thraktor. At least, Nintendo seems to have given this issue thoughts in their design. So I guess they are doing as much as we can expect from them.

And I am sorry for @-ing you @brainchild but based on the calculations from Thraktor, could a SoC that has a memory bandwidth of 88 GB/s run Nanite? We all know now that one would need speeds above 500 MB/s to stream data from the hard drive but can the bandwidth be a bottleneck?

I'll preface that question as usual by saying that I am a total noob in this domain. I am grateful for each answer I receive.
Well 88GB/s would be the ram bandwidth
Theoretically the actual throughput would be much faster because of the cache
It’s not quite as simple as it seems … that being said I don’t know what nanite needs
 
0
I've been out of the loop but are people expecting a new model to launch this year? I was assuming 2023 for some reason.
The only rumors we have heard have said late this year with a chance to slip into early next year.
 
0
So a YouTuber named RGD uploaded an informative and fascinating video about the hardware that was originally the successor to the Nintendo 3DS, but ultimately became the Nintendo Switch, called Project Indy.


One fact that intrigued me is that Nintendo actually considered using a 120 Hz display for Project Indy. That fact makes me believe that Nintendo going for a 1080p display with a refresh rate of 120 Hz is not as unlikely as originally thought, especially if Nintendo wants to enable VRR support for TV mode and handheld mode. (Granted I don't think the possibility is very high.)
 
Last edited:
120Hz would still be a crazy high power draw (and at 1080p? very high). not to mention the cpu and gpu costs
The only reason why I mentioned 1080p displays is because that seems to be the lowest resolution required to have VRR support on displays already available (e.g. iPhone 13 Pro, iPhone 13 Pro Max).

Of course, Nintendo could theoretically customise a 720p display to have VRR support, but I imagine that won't be cheap.
 
120Hz would still be a crazy high power draw (and at 1080p? very high). not to mention the cpu and gpu costs
Yeah the battery life differences are ridiculous on Steam Deck when going from 30fps to 60fps. Basically halves your battery life at 60fps. 120fps would be nuts.
 
0
Due to power draw reasons, I’d suspect that if Drake has support for 120Hz, it would be mainly a docked feature and for very select titles.
 
So a YouTube named RGD uploaded an informative and fascinating video about the hardware that was originally the successor to the Nintendo 3DS, but ultimately became the Nintendo Switch, called Project Indy.


One fact that intrigued me is that Nintendo actually considered using a 120 Hz display for Project Indy. That fact makes me believe that Nintendo going for a 1080p display with a refresh rate of 120 Hz is not as unlikely as originally thought, especially if Nintendo wants to enable VRR support for TV mode and handheld mode. (Granted I don't think the possibility is very high.)


When I see that list of display resolutions in the video ...it seems to me like those were display modes rather than completely different screens. though Its weird that 2 of those display modes are scratched out and are basically the same as the ones below... so I guess there could be some different screen implementations that were tested ...
For example if you used both displays at full resolution 854x480 each you were limited to 60fps each
OR similarly you could use the 3D display in stereoscopic 3D at half the horizontal res (just like 3DS did) and the second screen at full res at 60fps each
BUT if you only wanted to render using one screen at full res you could have the option of 120 fps


very interesting stuff.
 
0
The comforting thought for people worried about the price is that regardless of whether it'll be $499 or less, we won't be able to get our hands on it even if it did release 😁

😢
 
So a YouTube named RGD uploaded an informative and fascinating video about the hardware that was originally the successor to the Nintendo 3DS, but ultimately became the Nintendo Switch, called Project Indy.


One fact that intrigued me is that Nintendo actually considered using a 120 Hz display for Project Indy. That fact makes me believe that Nintendo going for a 1080p display with a refresh rate of 120 Hz is not as unlikely as originally thought, especially if Nintendo wants to enable VRR support for TV mode and handheld mode. (Granted I don't think the possibility is very high.)

I wonder... would people have prefered this Switch to the current one we ended up getting? What would the alternate timeline have looked like?

(The real question being "How many people here are actually against such 'gimmicks'?" I would presume the folks in Famiboard would be a bit more...open-minded.)
 
I wonder... would people have prefered this Switch to the current one we ended up getting? What would the alternate timeline have looked like?

(The real question being "How many people here are actually against such 'gimmicks'?" I would presume the folks in Famiboard would be a bit more...open-minded.)
We own a switch, a console with gimmicks attached to it and integral to its design, and we are here speculating on a switch but better, some even thinking of different ideas for the controllers, others as per the silicon of the switch, others about improving the overall experience, and you think most of us aren’t at the very least open-minded to other gimmicks? Lol

Personally, I’d like to see more VR/AR integration.
 
Quoted by: SiG
1
We own a switch, a console with gimmicks attached to it and integral to its design, and we are here speculating on a switch but better, some even thinking of different ideas for the controllers, others as per the silicon of the switch, others about improving the overall experience, and you think most of us aren’t at the very least open-minded to other gimmicks? Lol

Personally, I’d like to see more VR/AR integration.
You'd be surprised about how many people I've seen post here hate Nintendo to return to its "gimmick roots" as it were. Some people just want feature parity like the other two companies all the way down to copying achievements/trophies, to even naming the successor "Nintendo Switch Pro" or "Nintendo Switch 2" ;) .

Personally, I do want them to explore the Miracast/wireless streaming ideas more, though latency will be an ongoing issue, which might make the "docked" solution still viable (that and having a higher power profile).
 
Due to power draw reasons, I’d suspect that if Drake has support for 120Hz, it would be mainly a docked feature and for very select titles.
I think Nintendo (shall they go for a high-refresh panel) would probably be interested in a 40Hz mode for running their games with DLSS.
 
Ehh, I say moreso Nintendo's next finanical meeting covering FY23 would be the make or break for Drake 2022/2023

Now I'm still not entirely sold that it will release next FY.

But they dont really need to highlight it in any fashion when covering their upcoming FY, just like they didnt do with Lite, OLED or even the original Switch the FY the year it launched.
 
But the double-wide Orin tensor cores would be the more efficient option in performance per Watt if Nintendo's looking purely for DLSS performance. Let's say that you had a 6 SM part based on desktop Ampere and wanted to double the tensor core performance. If you switched to Orin's double-wide tensor cores, then you're approximately doubling the power consumption from tensor cores*, and that's about it. However, if you kept the standard tensor cores, but doubled up on the number of SMs, you're doubling the power consumption for everything. Tensor core power consumption doubles, because you've got two of them, but you're also adding extra standard CUDA cores, extra RT cores, extra texture units, extra control logic, and all the additional wiring, logic and associated power consumption from moving data and instructions to and between these units.

* I'd actually assume the power consumption of Orin's tensor cores is less than double the power consumption of standard Ampere tensor cores, as while you're doubling the ALU width, there'll be a certain proportion of instruction decode and control logic which won't be doubled.
So that says to me it's a cost consideration to go with the standard Ampere tensor cores.
In hardware design, performance, power efficiency and cost are the holy trinity of considerations, so if these double-wide tensor cores are more performant and more power efficient, the only other consideration left unaccounted for is cost. (size constraint is another factor specific to a hybrid, but I have doubts that these double-wide tensor cores you mentioned would drastically increase die size on the SoC)
My thinking would be that the boost in performance and power efficiency from these double-wide tensor cores come with a pretty hefty cost and Nintendo found more cost-efficient ways to boost performance and save power in lieu of better tensor cores, while still ensuring they have what they need to achieve what they're after for DLSS and anything else they intend to do with them.
Who is saying they are throwing away money?

All I’m saying is that any company spending 50% of their yearly revenue on new hardware R&D is taking a far FAR greater risk on invenstment than a company spending 5% of their yearly revenue on R&D. Doesn’t matter if the latter R&D amount is more than the former, the above is still true. This shouldn’t be a controversial statement.

One would also say, that if a company at one time was spending 50% of their revenue on R&D, and is now spending 5%, we would call that a company being far more conservative in investment spending, actually.

So, Nintendo spending on R&D for Drake is arguabley Nintendo being more conservative lol. They will actually waste less of their profits on Drake R&D than they did for Mariko.
See, this thinking inherently suggests that because it's a small percentage of revenue, they can afford to spend it regardless of return because it's a small percentage, but I can tell you with absolute certainty that Nintendo doesn't think this way, because they have first-hand experience with revenue being a moving target, something that can shrink drastically in a matter of a year, as you yourself acknowledge. Hell, most corporations don't operate this way. Project expenses aren't weighed against current revenues, they're weighed against the revenues and profits that project is projected to generate. Every dollar spent is a risk taken to make money, so for every dollar spent, they want that dollar back and more; anything else would be considered profit loss. A 10 billion yen R&D spending increase is still a 10 billion yen spending increase, regardless of how that's represented as a percentage of revenue. It's not more or less conservative based on what they raked in for the year. The only thing in question is whether those costs are properly returned with an appropriate increase to profits. And spending on what qualifies as next generation hybrid hardware development for iterative successor sales figures? That ain't wise spending.

To the last part that I bolded, though, you think that a brand-new custom SoC costs less money to design and/or provides a larger return on investment than a simple die shrink to a fully pre-designed SoC that reduced manufacturing costs of every Switch sold since July of 2019 (which is over half of all Switches sold and climbing)? You think Mariko was a comparative waste of money?
I'm sorry... WTF?! You must expect this iterative revision you imagine exists to double annual Switch sales figures for this to be taken seriously, because I couldn't stop laughing when I read that. I've tried writing this reply a few times now and had to stop each time because I'd start laughing again.
I think if that werent the case, and Nvidia was ok still making Volta architecture chips through 2027 and that was cost effective…we would have seen that. I’m guessing it isn’t. Heck, people here have even intimated the idea that maybe fewer SM’s and 8nm might not have been enough to function overall well enough.

We can agree to disagree, but I’m positive the hardware usage that they ended up with was more out of necessity to get everything DLSS related working adequately with the caveats Nintendo creates for themselves with Switch hardware.
Nvidia has been more than OK to still be making Maxwell architecture chips through to 2023 (or 2027 as you suggest, if we're to believe that this is an iterative revision released alongside current Switch models to elongate its hardware cycle). That'd be 8 (or 12, again as you suggest) years from when the Tegra X1 was first commercially used.
But NOW suddenly they're going to take issue with making chips on old architectures for another 4 years until you believe a true successor will happen? C'mon now, that ship has long since sailed.
Look, the RTX 2060 is the lowest CUDA offering by Nvidia to attempt DLSS. That’s 1920 CUDA running at 1.37 ghz. and draws 175W.

That can play a 2019 game at 4K/60fps…but only in performance mode. Starts to suffer beyond that.

And you want me to think Drakes 1536 CUDA at 1GHz with a 25W power draw is someone bordering on grotesque overkill to play games at 4K/60fps? Why?

Go look at the 3050ti DLSS capable laptop gpu. That’s supremely grotesque compared to Drake lol

What’s the minimum amount of CUDA cores running at 1 ghz with 25W is needed to get…say…Death Stranding running at 4K/DLSS in performance mode? I genuinely don’t know.

As I mentioned above, the RTX 2060 barely does this.
We’re talking about hardware to render Switch games in 4K and you bring up… a PS4/PS5/PC exclusive? Huh? What a red herring you decided to throw into the conversation. How about we look into what’s required to achieve upscaling with Switch games themselves, hmm?
Xenoblade 2 on PC via Yuzu in double its native resolution with minimal to no visual glitches only needs… a GTX 1050 Ti, a budget Pascal GPU (read: pre-DLSS). Most of the videos and info I'm reading say, for ideal performance at UHD 4K and 60fps, a mid-range Turing GPU is able to tackle that pretty handily for the games I was looking at (and a Turing GPU-based SoC is also a cheaper and capable option in lieu of an SoC with an Ampere GPU). Keep in mind, these cards can achieve these results using unofficial emulation and inefficiently-optimized software modding on PC.
So, since you've made the case that this is an iterative revision with likely few to no exclusives intended primarily to extend the Switch's lifespan to 2027, it doesn't need to achieve more than that, and no data we have at our disposal about what's required to get Switch games to reach 4K (and 60fps, no less) suggests that an Ampere GPU SoC is at all necessary.
The proof that Nintendo/Nvidia couldn’t make a cheaper, more efficient SoC to get 4K/60fps DLSS gaming in is in the fact that it never materialized. If there were such a possibility, it would have happened.
Unless... y'know, what you think they're designing (an iterative revision) isn't actually what they're designing and that's why it didn't materialize. So the absence of evidence is not really "proof", is it?
shrugs you say demand, I say incentivize.

I’m pretty sure Nvidia greatly informed Nintendo their best options to get Swifch mobile DLSS gaming workable and sustainable for 2022-2027 or so.

Nintendo went for the overall best option. They can afford to spend a bit more on a model to future proof it better. Their current success and the knowledge that investing in this same ecosystem is important to maintain that.

I’m sure Nvidia convinced Nintendo on the value of DLSS.

They weren’t bamboozled by Nvidia on anything lol
I'm sure Nintendo saw the value in DLSS as well, but Ampere isn't required to achieve it. And since you think this hardware is only to last 4 years and will not be sustaining its own software library as an iterative revision to be replaced by a successor in 2027, there's no "future-proofing" required. Saying it needs to be future-proofed is actually working against your own argument that this is merely a 4-year stopgap.
 
Now I'm still not entirely sold that it will release next FY.

But they dont really need to highlight it in any fashion when covering their upcoming FY, just like they didnt do with Lite, OLED or even the original Switch the FY the year it launched.
I think Alovon means we night get a sense of whether they're releasing new hardware at that meeting based on the forecast, or better yet if someone asks Furukawa and he says "we cannot announce anything at this time" or similar.
 
0
Now I'm still not entirely sold that it will release next FY.

But they dont really need to highlight it in any fashion when covering their upcoming FY, just like they didnt do with Lite, OLED or even the original Switch the FY the year it launched.
Nobody said they have to highlight. We see how Furukawa responds to investors. His phrasing of words will make it clear if it’s coming. Idc what anyone else says other than him. It’s his product he knows when it’s coming and he’ll answer for it in a roundabout way.
 
0
Personally, I do want them to explore the Miracast/wireless streaming ideas more, though latency will be an ongoing issue, which might make the "docked" solution still viable (that and having a higher power profile).
Same.

I'm simply speculating here, but maybe Miracast or a similar feature could be used for a Nintendo Switch Lite model equipped with Drake as the SoC, but with the max resolution limited to the max resolution of the display, for battery life and latency reasons.

 
So a YouTube named RGD uploaded an informative and fascinating video about the hardware that was originally the successor to the Nintendo 3DS, but ultimately became the Nintendo Switch, called Project Indy.


One fact that intrigued me is that Nintendo actually considered using a 120 Hz display for Project Indy. That fact makes me believe that Nintendo going for a 1080p display with a refresh rate of 120 Hz is not as unlikely as originally thought, especially if Nintendo wants to enable VRR support for TV mode and handheld mode. (Granted I don't think the possibility is very high.)

Uh, why is that the thumbnail?
 
Uh, why is that the thumbnail?
I guess because of this.
sw-leak-2.jpg



In addition, Apple will need to increase memory bandwidth for their SOC. They have been able to stay on LPDDR4x for 5 generations by increasing on die cache sizes and improving utilization. With the A16, the step up to LPDDR5 brings a large cost increase. Due to the way the 3 major DRAM companies have slowly increased output, cost per bit of DRAM has not really fallen. Memory prices are a major drag on computing.
Memory-1.png

The iPhone 13 ships with 4GB of LPDDR4x, moving to 6GB of LPDDR5 and to the larger A16 chipset would be cost prohibitive. This would result in more than a $40 increase in bill of materials (BOM). With numerous supply chain disruptions and cost increases for energy and commodities, it would be a difficult pill for Apple to also swallow the ballooning silicon costs as well. Apple could also increase prices, but that likely would eat into sales volumes.
 
I wonder... would people have prefered this Switch to the current one we ended up getting? What would the alternate timeline have looked like?

(The real question being "How many people here are actually against such 'gimmicks'?" I would presume the folks in Famiboard would be a bit more...open-minded.)
Maybe Nintendo fans would have liked it but I doubt it would be the sucess Switch is.
Uh, why is that the thumbnail?
Because he infer from the leaked docs and designs that this Indy console would have the design of that infamous patent, which was used for the fake NX rumor.
I guess because of this.
sw-leak-2.jpg
I would have loved this design. Seems like more ergonomic than current Switch. Maybe a redesigned Switch Lite down the line.
 
0
You'd be surprised about how many people I've seen post here hate Nintendo to return to its "gimmick roots" as it were. Some people just want feature parity like the other two companies all the way down to copying achievements/trophies, to even naming the successor "Nintendo Switch Pro" or "Nintendo Switch 2" ;) .

Personally, I do want them to explore the Miracast/wireless streaming ideas more, though latency will be an ongoing issue, which might make the "docked" solution still viable (that and having a higher power profile).
Same.

I'm simply speculating here, but maybe Miracast or a similar feature could be used for a Nintendo Switch Lite model equipped with Drake as the SoC, but with the max resolution limited to the max resolution of the display, for battery life and latency reasons.
mmm, maybe:


Not what you have in mind, but they seem to want to do something here.
 
Since we sometimes talk about cloud gaming here, this recently published Sony patent may be of interest:


It describes a cloud gaming server that sends a lower-res video stream and an “additional data” stream to the client; the client performs a super-resolution operation by combining these two streams. The “additional data” may include:

“an edge strength of the original image, a depth of each pixel included in the original image, a texture type of the original image, an optical flow of the original image, and information representing a direction and a velocity in which and at which a rectangular region in the image moves.”

0y7VXhtAXrZPFsHtEziFQZZKKsuknEukANuovlckTChWyZeHOAfwtMfqQKrY7lsBDl_Kdgwlq7z_YkZT-yjvUwKeOhyMLuJwQIy6-o7D6J7mYdjDbs-YmIwLWpaz2xhhs72aKjsrGpazpnQPep1qdgIqeJoxjlVc6Oii40s2WRTCxe-HAxIawCSnwHke2oEC
 
Personally I'm not too perturbed by the prospect of a premium Drake Switch being priced at $450+ (and perhaps a dock/accessory-less SKU for $400). It's actually a good product strategy to target more market segments via price/product differentiations.

What baffles me, however, is that the current Lite and hybrid models may not have enough legs to last for more than a couple more years, and sooner than later they will need to be upgraded to the new platform standard (Drake SoC, wider memory, faster storage, etc.) while ideally maintaining their respective price points at $200 and $300-$350. Looking at the rather high-end specs being speculated in this thread, I'm not sure how that could be accomplished.

Even if we humor this idea that Nintendo somehow manages to release a Drake-capable Lite and hybrid for $200-$350, it'd destroy the value of the $450 premium Switch and alienate their most loyal early adopters (owners of the $450 Drake).

A potential solution to this conundrum is to double down on the price/product differentiations by introducing two Switch tiers a la Xbox Series X and S. But unlike MS who released both tiers simultaneously, Nintendo would introduce the premium tier (let's call it Switch DX for now) first to capture the enthusiast/early adopter segment, and the standard tier 2-3 years later for the mainstream/casual markets.

Switch DXSwitch SSwitch Lite S
2022-20232024-20252025-2026
Drake (12 SM, 8 A78), 128-256GBDane (6 SM, 4-6 A78), 64GBBinned Dane, 32GB
Hybrid + enthusiast features (e.g., live streaming)HybridHandheld only

IMO the staggered release of tiered products is a superior strategy because it not only allows Nintendo to extract more value from the higher-end segment sooner, but also drives down production costs and builds a larger software catalog in advance of the lower tier launch.

While this speculated lineup may seem unlikely, it isn't impossible either for the following reasons:
  • The mysteriously disappeared Dane may make a triumphant return
  • Orin ADAS (5w-10w) and Jetson Nano Next are being released in 2023; could be related to Dane
  • The software compatible layer developed for Drake to run TX1 titles might simplify the support for yet another SoC (Dane)
  • Cross-gen games that run on both TX1 and Drake should be able to support Dane easily
  • Drake exclusive games already give up on the 100 million TX1 installed base; it's no great loss if they aren't patched to support Dane
If this model works out, Nintendo may sustain it by introducing the Switch 2DX in, say, 2027-2028 with a new SoC and new gimmicks; 2-3 years later, the Switch 2S and Lite 2S may follow with a die-shrunk/refreshed Drake. Rinse and repeat for a tick-tock iterative succession plan.

Edit: typo

I think you're underestimating the headache of game support between these devices, particularly if they're not released at the same time. While cross-gen games could run on the Switch S, they would by default have to run at original Switch settings, and there's no guarantee that they would get patched. Combined with simply not running games exclusive for Switch DX (which would probably be most big third party titles), it would potentially be quite disappointing for anyone expecting an Xbox Series S style experience (where it runs every Series X game, and always with significant improvements over Xbox One versions). It would also be a pain for developers, as anyone intending to support full cross gen would have to support six different profiles (handheld and docked for each of Switch, Switch DX and Switch S), and even going forward they'd need to support four profiles compared to the current two. I'd imagine third party developers who released Switch DX exclusives would also be pretty disappointed to discover that their games won't run on Nintendo's new device.

The way I could see this working is if Nintendo release Switch DX with a 1080p screen and choose to disable some SMs (let's say reduced to 6 active SMs) in portable mode. They could then design a cut down version of Drake with only 6 SMs, and use that for Switch S, which would use Switch DX's portable mode in both docked and portable setups. They would still have to have the same number of CPU cores, the same amount of RAM, etc., which would limit cost reductions, but they could cut storage, use a lower quality screen, etc. to bring down the price. Software support would be far simpler, as games would only need two profiles to cover both devices, and all Switch DX games should run on Switch S (with significantly enhanced performance over the base Switch) without any issues.

The other aspect of Microsoft's strategy with the Series S, though, is that they want to maximise the growth of the Xbox Series platform as a whole, and sell devices which they will continue seeing revenue from for a much longer period. If they were trying to maximise short-term profit, then they would just continue selling Xbox One S consoles instead, but someone buying an Xbox One in 2021 only going to be buying new releases for maybe 2 years (or paying for Gamepass for a similar time), whereas someone buying an Xbox Series S will be getting new releases for perhaps 7 years (and ideally paying for Gamepass for that entire time). Hence Microsoft's willingness to take a loss on the Series S vs just continuing to sell the Xbox One S.

Of course Nintendo don't have anywhere near the proclivity for loss-making on hardware as Microsoft do, but the basic logic still applies. A new customer buying the Switch DX is much more valuable to them than a new customer buying the base Switch in 2022/2023. There's an assumption in a lot of the discussion around price here that Nintendo will want to make even more of a profit off the Drake Switch than they do off the original model, but I don't necessarily agree with that. Looking at the hardware it would be very surprising to me at this point if this isn't functionally the Switch 2, and that being the case they will have a strong incentive to sell Drake Switch over the base models, as that will make them more profit in the long run.

I definitely don't expect them to sell it at a loss, but I could definitely see a much tighter margin than we've got on the base Switch at the moment.

So that says to me it's a cost consideration to go with the standard Ampere tensor cores.
In hardware design, performance, power efficiency and cost are the holy trinity of considerations, so if these double-wide tensor cores are more performant and more power efficient, the only other consideration left unaccounted for is cost. (size constraint is another factor specific to a hybrid, but I have doubts that these double-wide tensor cores you mentioned would drastically increase die size on the SoC)
My thinking would be that the boost in performance and power efficiency from these double-wide tensor cores come with a pretty hefty cost and Nintendo found more cost-efficient ways to boost performance and save power in lieu of better tensor cores, while still ensuring they have what they need to achieve what they're after for DLSS and anything else they intend to do with them.

There is certainly a cost to going with larger tensor cores, which should largely just be a reflection of the larger die area, but there's also similarly a cost to adding more SMs. I don't disagree that performance, power efficiency and cost are what Nintendo and Nvidia would be optimising around (or PPA, performance, power and area, as ARM refer to for their mid-size cores like the A78), but I think the GPU they've ended up with can tell us quite a bit about what kind of performance they were optimising around. In particular, if their main intention was to run base Switch games in 4K using DLSS, then the PPA-optimised design probably would have had less raw GPU performance and more of a focus on tensor core performance, ie something like 4 or 6 SMs with bigger tensor cores, because they don't need to render games at more than around 1080p base, and can let DLSS do the heavy lifting.

Instead, we got a much bigger GPU with more of a balance between raw GPU performance and tensor core performance. This is going to be more expensive and power-hungry than the kind of design I was expecting, which tells me their performance expectations must be much higher than just running base Switch games in higher resolutions. To me it's more in-line with a PPA-optimised design for an entirely new console, rather than a 4K version of the base Switch.
 
I guess because of this.


So, is too expensive for Apple to use LPDDR5/x but it is widely used by Android phones? Is this because the ultra wide Apple architecture?
 
Last edited:
I really hope we get some new information coming out of the GDC, while the leak was fun it stills feel like there is so many other missing components
Let’s not get greedy. If you had told us the details we got were coming a month ago we’d have said we’d have been happy until the official reveal 😝
 
The Switch Lite is a “damaged good” in order to lower the price point. Equipping it with the video streaming capability (wireless docking) would defeat the purpose. See this Caltech paper on Pricing Damaged Goods:

“Companies with market power occasionally engage in intentional quality reduction of a portion of their output as a means of offering two qualities of goods for the purpose of price discrimination, even absent a cost saving. This paper provides an exact characterization in terms of marginal revenues of when such a strategy is profitable, which, remarkably, does not depend on the distribution of customer valuations, but only on the value of the damaged product relative to the undamaged product. In particular, when the damaged product provides a constant proportion of the value of the full product, selling a damaged good is unprofitable.”

As for the hybrid model, wireless docking isn’t really necessary, since the user can dock physically near the TV and use a wireless controller away from TV. The main potential benefit for the hybrid model to wireless dock is to facilitate asymmetrical gameplays; the Wii U, unfortunately, demonstrated that very few games could do that compellingly and not enough players cared about them either.

The video streaming capability would be more interesting, IMHO, in enabling a more powerful Switch model (Drake?) to host party gameplays by streaming to multiple other Switches or even phones (running the NSO app) without requiring the other devices to download the same game. It’d encourage a household to purchase multiple Switch units, and casual players to download the NSO app and create Nintendo accounts.
 
Just to frame exceptations better, what preqreuisities must be fulfilled (in terms of clock speeds, cache, node and RAM) for Drake to be as powerful as a GTX 960/GTX 1050Ti, and is there a realistic scenario in which it can match their rasterization performance?

For the record, here are the specs:

GeForce GTX 960
TSMC 28 nm
Die size 398mm2
1280 cores
Base clock core 924 MHz
4096 MB RAM GDDR5
120 GB/s bandwidth
Bus width 192 bit
L2 cache size 1.5 MB
2.3 TFlops
145 W

GeForce GTX 1050 Ti
Samsung 14 Nm FinFET
Die size 132mm2
768 cores (6 SM)
Base core clock 1290 MHz
4096 MB RAM GDDR5
Bus width 128 bit
L2 cache size 1MB
1.9 TFlops
75 W

If Drake is anything like these, then we are in for a ride. A good one I might add.

My laptop has a GTX 1050 and Yakuza Like a Dragon barely moves at 900p, average quality and even pulling down in some respects and barely 30fps, I really expect more performance than that for "Switch Next Gen/2/Pro". The numbers say one thing, but when I play Yakuza Like a Dragon on my laptop it feels like the graphics quality of my laptop for Yakuza 7 could be a Switch 2017 version of the game.
 
0
Do we usually get any hardware related news at GDC? Also, has Bloomberg stated anything after the Nvidia leak?
Depends on when Nintendo plans on launching new hardware for the first question. And no for the second question.
 
Do we usually get any hardware related news at GDC? Also, has Bloomberg stated anything after the Nvidia leak?
GDC is in person this year, so it's gonna be the first time in a while where a lot of developers come and talk/share rumors and such. We're not expecting anything official, but if something is really releasing within the next 12 months it would be very surprising if we don't hear any rumblings about it shortly after GDC.

As for Bloomberg, I don't think most major publications will touch illegally leaked data with a 10 foot pole.
 
the GDC after Drake's launch could be interesting if Nintendo or affiliates talk about optimizing for new hardware. I think Nvidia will host some talks about leveraging hardware with a developer, like they did for Switch (Witcher 3 and Sinking City)
 
Note actual news, but it is generally a pretty ideal environment for spreading rumors around.
Ah, this makes sense
GDC is in person this year, so it's gonna be the first time in a while where a lot of developers come and talk/share rumors and such. We're not expecting anything official, but if something is really releasing within the next 12 months it would be very surprising if we don't hear any rumblings about it shortly after GDC.

As for Bloomberg, I don't think most major publications will touch illegally leaked data with a 10 foot pole.
Oh! I see, hopefully we do get something, even if small. I'd love it if this Switch launches end 2022 beginnings of 2023, I do need another Switch but I'd like to know if there is possibility of the new one launching soon so I save for that one instead. And yeah, you are totally correct, I forgot that it was a ransomware attack for a moment there.
 
0
I wonder... would people have prefered this Switch to the current one we ended up getting? What would the alternate timeline have looked like?

(The real question being "How many people here are actually against such 'gimmicks'?" I would presume the folks in Famiboard would be a bit more...open-minded.)
the fact that this thing used a modified Wii U gpu...I imagine if the Wii U was more successful we could have seen a timeline where they built off of that ecosystem... sharing a GPU arch would have sort of been a step in that "unifying development" they were undergoing at the time. Pretty fascinating to think about.
I imagine they might have stuck with AMD and I bet the console and portable would have played together more
 
0
So, is too expensive for Apple to use LPDDR5/x but it is widely used by Android phones? Is this because the ultra wide Apple architecture?

I don't think it's "too expensive", but their recent SoCs have gone in the direction of very large caches. This improves performance and power efficiency, and it has the side-effect that they don't really need faster memory, so they can save money by going with LPDDR4X.
 
GDC is in person this year, so it's gonna be the first time in a while where a lot of developers come and talk/share rumors and such. We're not expecting anything official, but if something is really releasing within the next 12 months it would be very surprising if we don't hear any rumblings about it shortly after GDC.

As for Bloomberg, I don't think most major publications will touch illegally leaked data with a 10 foot pole.
Ars Technica reported on the leak and the demands of the leakers.

In depth reporting on the leak for Nintendo has been scant, though Kotaku was on strike when it occurred and DF touched on it, but as this thread is pointing out, the leak confirms many things about the hardware in development for the "Next Switch". However, power consumption affects much of how to interpret the leak for comparison to other hardware as well as its ability to drive 4K. Also its potential BOM cost is weirdly out of line with previous Nintendo hardware.

Personally I love seeing the work @Thraktor and others are doing on the topic, it will be interesting to see what comes from GDC.

In either event, I would expect supply chain leaks some point soon if a late 2022 launch was still on the table.
 
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