• 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.
  • Do you have audio editing experience and want to help out with the Famiboards Discussion Club Podcast? If so, we're looking for help and would love to have you on the team! Just let us know in the Podcast Thread if you are interested!

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

Or perhaps Orin Nano was expected earlier in the year, as production samples failed to hit yields. Orin Nano 4GB is an excellent base for a mid-gen Switch, if you ignore timing. The power consumption is right in line with the base Switch, it's got 2x the SMs as Mariko, and 6 CPU cores instead of 4. You could get Mariko level batter life with a little DLSS, or extended power across the board, for a cost about the same as the existing device.
Nothing with DLSS and RT cores and, most of all, a lack of native BC, is a believable chip for a mid-gen refresh. I also think this post is far too easily connecting the dots between three different people's reporting (or offhand comment in DF's case) in order to come up with a clean narrative. @Skittzo above has some of the same thoughts I do on this.

In a vacuum: If you remove the context of the timing and reporting and everything else, I could believe that some other chip was on the table for Nintendo, maybe a binned Orin like people speculated way back when. But it's just not possible for that hardware to have been a mid-gen refresh, or for it to be cancelled. At most, it evolved into a new chip with the same features and therefore likely the same positioning and release window as originally intended. Not in a vacuum, even this narrative has to contend with the fact that whatever evolutions or decisions were made around hardware would have happened before third parties knew about any of these plans. I agree with much of what @Thraktor said about this in his previous post.

Hidden content is only available for registered users. Sharing it outside of Famiboards is subject to moderation.
I'm not an uncle believer myself, but wasn't the supposed January window based on their response (~6 months IIRC) when someone asked how long prototyping to production usually took?
 
I'd say I have multiple reasons to suspect that this is the case, given to me by 3 different sources, but because I can't talk about any of those here, I'll just say that I'm PRETTY SURE that I was right a year and a half ago when I said that Drake probably moved from 8nm to 5nm. Those who need to think I'm an insider, I'm not, but I've always shared insider knowledge that I come across, this is my big wink, I don't normally do this, and I have 3 separate sources, they could be lying, but 2 of them have been very reliable to me in the past and the other is someone I've known and talked with on this forum for years and have no reason to think they are lying... 2 pieces of this information came to me yesterday, the other I heard a few months ago... regardless I'm confident that Drake was retooled in 2021 and in early 2022 the engineering sample for Drake on a more advanced node was released, and is now ready for full scale production. I haven't heard any updates on 1H Drake release beyond what has been shared here though.
We also have this article from earlier in the year, stating that Samsung claimed 5nm yields increased and this report in September stating their 5nm yields could be as high as 70% (not sure how good or bad that is, maybe it'll be even better by May?).
* Hidden text: cannot be quoted. *
So nothing explicit, but some pretty strong implications it seems like?
 
we know NVN2 on windows is for SDK development, the test only makes sense as estimations of T239, we also know there is no RTX GPU that can run those clocks at that power draw, there is definitely a lot of deductive reasoning going on with my statements, I admit to that, but literally that test cannot make sense on any current RTX GPU, it only makes sense on Drake if Drake is twice as efficient as Ampere, unless you know of an Ampere GPU with 6SM right?

I think my posts come across to people clearly, it's just that some don't like how conclusive I'm being, that is my opinion though, if I'm missing something in my deductive reasoning, pointing it out can be helpful, but I don't believe that information exists without it being a tool for Drake. There was also someone, maybe oldpuck who said that that test is definitely Drake, I don't really need anyone to say this, because I concluded as much. Drake is also the only NVN2 GPU right? so how is this test unrelated to it? What substitute could be put forward that would only draw 9.3w at 1.125GHz? It's also right in line with OG Switch power draw, so yeah it seems very unlikely that those numbers aren't targets for Drake, why would they test 3 different numbers if it wasn't for Drake?

I think the problem is that you believe it can't be used to make concrete conclusions, but the test's existence actually does call into question what information does it provide, and in that we can make reasonable conclusions, and that is what I've done here, you found great evidence, you can hold your position, I'm not trying to tell you that you have to have mine, but that information is now part of the discussion around Drake and that is absolutely reasonable. We have no reason to believe it is an Ada GPU used in this test from last year, we know it was done in NVN2, and that no Ampere GPU can deliver those performance numbers, so obviously it's reasonable to conclude that it is Drake, and further, those power numbers exist, the worst case is that it is just for the GPU, and that is what I've gone with. I respect that a lot of posters here want to play conservatively with speculation, but that isn't my game, my game is deductive reasoning, and anyone with a brain who isn't trying to avoid being a bit reckless maybe, should conclude that is Drake in those tests and that Drake likely shrunk...

I'd say I have multiple reasons to suspect that this is the case, given to me by 3 different sources, but because I can't talk about any of those here, I'll just say that I'm PRETTY SURE that I was right a year and a half ago when I said that Drake probably moved from 8nm to 5nm. Those who need to think I'm an insider, I'm not, but I've always shared insider knowledge that I come across, this is my big wink, I don't normally do this, and I have 3 separate sources, they could be lying, but 2 of them have been very reliable to me in the past and the other is someone I've known and talked with on this forum for years and have no reason to think they are lying... 2 pieces of this information came to me yesterday, the other I heard a few months ago... regardless I'm confident that Drake was retooled in 2021 and in early 2022 the engineering sample for Drake on a more advanced node was released, and is now ready for full scale production. I haven't heard any updates on 1H Drake release beyond what has been shared here though.
I don't know what else to say.

You having your reasons to believe in 5nm or whatever is fine. What I posted is not evidence of that or of clock speed targets. It's just evidence that Nvidia was testing power consumption and related metrics for certain DLSS profiles in NVN2. The specific numbers they used are not conclusive of anything, because the tests weren't done on T239, so the clock speeds they chose to set were being set on some unknown Windows hardware, not T239, and therefore the numbers can't be assumed to be relevant to T239.
 
I don't know what else to say.

You having your reasons to believe in 5nm or whatever is fine. What I posted is not evidence of that or of clock speed targets. It's just evidence that Nvidia was testing power consumption and related metrics for certain DLSS profiles in NVN2. The specific numbers they used are not conclusive of anything, because the tests weren't done on T239, so the clock speeds they chose to set were being set on some unknown Windows hardware, not T239, and therefore the numbers can't be assumed to be relevant to T239.
Correct me if I'm wrong, but those SDK settings, like the original hacked leak of Switch's SDK, emulate a Switch environment for development. This happened in the SDK for Switch when you set for handheld or dock modes, a test done in this SDK would only work that way... If they weren't emulating a Drake environment why would they even jump from 660MHz to 1.125GHz to 1.38GHz? The reasonings I give are clear, you make it sound unreasonable, but that isn't how the previous SDK worked, at least not at launch, I don't have access to a current Switch SDK, but pretty sure NVN on Windows just emulates Switch hardware still so you can make Switch games.
 
Correct me if I'm wrong, but those SDK settings, like the original hacked leak of Switch's SDK, emulate a Switch environment for development. This happened in the SDK for Switch when you set for handheld or dock modes, a test done in this SDK would only work that way... If they weren't emulating a Drake environment why would they even jump from 660MHz to 1.125GHz to 1.38GHz? The reasonings I give are clear, you make it sound unreasonable, but that isn't how the previous SDK worked, at least not at launch, I don't have access to a current Switch SDK, but pretty sure NVN on Windows just emulates Switch hardware still so you can make Switch games.
This seems to be conflating a few different things, like Nvidia's NVN2 source versus the Switch SDK, and the difference between a Nintendo SDK-defined performance configuration and the loose numbers in Nvidia's Windows-only test scripts that my post was about. And again I'm not sure how to respond beyond what I've already said.

I think I was very thorough in explaining the context (or lack thereof) for the numbers in my original post. If you don't believe me that these numbers aren't conclusive of anything, despite all that, then I can't convince you, I suppose. But I will still try to keep people informed if I see claims about TFLOPS being made based on what I posted.
 
Here's the thing though: John didn't say anything about the cancelled revision being DLSS capable. The only person who has said that is Nate.

There's a very simple reason to explain Mochizuki's reporting too, it's the same reason we've had for over a year. You hear from the manufacturing contacts that a new console is being assembled, you hear from the development contacts that new hardware is being developed for.

I don't buy that conflation. Developers said that their software was targeting a 2022-2023 window. Developers, not manufacturers, folks who would be the source for the 4k/DLSS devkit info. Unless there were multiple devkits in flight, the devkits in 2021 were not for a TX1 based piece of hardware, and those devkits were targeting this window for release.
 
I don't buy that conflation. Developers said that their software was targeting a 2022-2023 window. Developers, not manufacturers, folks who would be the source for the 4k/DLSS devkit info. Unless there were multiple devkits in flight, the devkits in 2021 were not for a TX1 based piece of hardware, and those devkits were targeting this window for release.
That window has not passed though... Mochizuki was claiming it would launch in 2021, it's hard to resolve that with the idea that a planned 2022/2023 launch was canned.

Conflation is really the only explanation for how he thought it would launch in 2021 and would be revealed "imminently" around E3.
 
We know that most likely what happened in 2021 was a mix-up between sources reporting on the OLED Switch and a Switch PRO, I can totally see that happening again, with a mix of sources talking about the cancellation of the refresh and what was supposed to be Drake
 
0
That window has not passed though... Mochizuki was claiming it would launch in 2021, it's hard to resolve that with the idea that a planned 2022/2023 launch was canned.
He was not? I am not sure if you're referring to Mochizuki's earlier articles, I'm referring to his followup after E3, in september of last year. Slightly edited:

By the time Nintendo showed off [the OLED model] in July, the company had already handed out the 4K kits to outside developers and asked them to design software to support the higher resolution... Developers declined to speculate on Nintendo’s plans for another console but said they expect to release their 4K Switch games during or after the second half of next year.
In late 2021, Mochizuki was told by developers that they planned on releasing 4K Nintendo software around this time. This is not a conflation between manufacturing and developers. After Mochizuki made the conflation mistake earlier that year, he was explicit to say his sources said 4k switch games in this window. The launch window and the 4k number are the same source, the developers making the game.
 
He was not? I am not sure if you're referring to Mochizuki's earlier articles, I'm referring to his followup after E3, in september of last year. Slightly edited:


In late 2021, Mochizuki was told by developers that they planned on releasing 4K Nintendo software around this time. This is not a conflation between manufacturing and developers. After Mochizuki made the conflation mistake earlier that year, he was explicit to say his sources said 4k switch games in this window. The launch window and the 4k number are the same source, the developers making the game.
Oh then we're misunderstanding each other, I was 100% referring to the mishap earlier that summer. I thought you were too.


I'm still not understanding how this supports the idea that a Drake revision was cancelled though. We're still in the window of late 2022-2023.
 
t239 exists without a doubt. The Nvidia leak links the t239 to Nintendo without a doubt.

But how complete was the NVN2 implementation? Are we really sure it's meant for Nintendo's next hardware?

I heard about a developer who pitched the X1 to Nintendo years ago, I think I heard it here, some hundred pages ago. Nintendo was supposedly inclined to use the Tegra K1 before that pitch. I'm guessing there is probably a proof-of-concept implementation of NVN targeting the K1 somewhere.

Could the NVN2 implementation targeting the t239 be just a proof-of-concept that was pitched to Nintendo, but ultimately declined?

The t239 does seem like a big fu**** chip. Maybe its purpose was Nvidia's attempt to enter the gaming laptop/chromebook market. It would explain the discrepancies around its power envelope.

We would be back to square one.

I don't think so, but I believe the probability is >0%. Next week's CES may bring some answers.

Anyway, assuming the t239 is Nintendo's, I don't think it's crazy they may sit on a finished ship for one year, to mature the OS and drivers, schedule some games, properly setup manufacturing, have an extensive inventory, etc. The Switch released in 2017 with a chip that released in 2015, but I know it's not exactly comparable.

I don't think so. I would sell early, start building momentum, and throttle production according to demand.

Anyway (again), Nintendo has been too weird lately. We know about a title in January, then one in February, then another in May which we know very little about. Nothing between February and May. That's weird from a company the publishes a dozen games every year, specially with the Mario Movie releasing in April. Then there's Pikmin 4 which had a very odd presentation and no gameplay, which (again) is odd for a game releasing in less that one year.

They are either up to something, or they're trolling us big time.

Best case scenario:
We haven't heard about the Mario team in a long time ... A new 3D Mario releasing in April with the t239 Switch 2 (same day as the movie with ads on every cinema screen).

Middle case scenario:
Switch 2 releasing holiday 2023 with one exclusive and some performance patches for existing games.

Worst case scenario:
Nintendo is panicking, declined the t239, is gonna ride the current model until the death of the universe.

I'll be fine either way :) Thank you all for the entertainment.

ETA: It's funny how the Mario Movie was delayed to the empty timeframe and Zelda delayed to a little after that.
 
Last edited:
But how complete was the NVN2 implementation? Are we really sure it's meant for Nintendo's next hardware?
this is impossible to know since we have no access to documentation. and what else is NVN2 supposed to be for? it directly references ray tracing and dlss

Could the NVN2 implementation targeting the t239 be just a proof-of-concept that was pitched to Nintendo, but ultimately declined?
there's no reason to believe this

The t239 does seem like a big fu**** chip. Maybe its purpose was Nvidia's attempt to enter the gaming laptop/chromebook market. It would explain the discrepancies around its power envelope.
we don't know how big it is, we can only assume. could the T239 make it to laptops? sure, but again, we don't know what it's for beyond being made for nintendo. remember there is also Orin which is more powerful and can also be put into laptops
 
0
Look at it from Nintendo's perspective: Having the majority of your software at soon after launch carried by third party software hasn't worked well for them in the recent past. Wii U and especially 3DS had "light" launches with 1st party games like another NSMB and Nintendoland or Nintendogs, Steeldiver, and Pilotwings. No wonder there was no hype at launch. Recent Nintendo fans (at the time) had no real attachment to many third party franchises due to their absence on N64 forward or the handheld support being mostly spin-offs or late ports.

When the 3DS launched the biggest 3P launch title was arguably SSF43DE and I didn't get that. I picked up pilotwings half a year later with a 50% bundle deal at gamestop with SF643D. I literally didn't buy any 3DS games at launch because nothing excited me on the 1st party side and I was unfamiliar with nearly all the third party games (not to mention there were so many to choose from at launch). The launch ended up so bad that Nintendo slashed the price within months.

The Wii U...well I got the deluxe bundle and I bought NSMBU...and that and Nintendoland were fine. But that was it really and again, I was completely unfamiliar with all the 16 other 3P games due to no prior experience on GBA, DS, and Wii. And the launch of that system was so bad that Nintendo had that infamous "panic" January 2013 Direct where they laid all there Wii U cards on the table.

And then the Switch exploded out the gate despite being a "Zelda machine" at launch. But that proved to not necessarily be a bad thing. At the end of the day, a single piece of high quality software AND the promise of heavy hitters in the near future and beyond was able to sell the system. Like, yeah it was just Zelda for 2 months, but people knew MK8D was coming soon, Minecraft was after that, Splatoon 2, M+R, "Pokemon" with Pokken, FIFA/NBA, FEWarriors, SMO, XB2. I don't think 2023 will quite be like that for Switch 2, but they'll likely follow a similar strategy (but through enhanced games, not 1P exclusives).

So TotK at launch, Pikmin 4 summer, some HD ports/remasters, DK?, 2D or 3D Mario, a new IP?, DLC, and other various games. I think one of the first big 3P games could be SF6. Then you announce real, modern FIFA, Madden, and everything else to round out the year.
 
Look at it from Nintendo's perspective: Having the majority of your software at soon after launch carried by third party software hasn't worked well for them in the recent past. Wii U and especially 3DS had "light" launches with 1st party games like another NSMB and Nintendoland or Nintendogs, Steeldiver, and Pilotwings. No wonder there was no hype at launch. Recent Nintendo fans (at the time) had no real attachment to many third party franchises due to their absence on N64 forward or the handheld support being mostly spin-offs or late ports.

When the 3DS launched the biggest 3P launch title was arguably SSF43DE and I didn't get that. I picked up pilotwings half a year later with a 50% bundle deal at gamestop with SF643D. I literally didn't buy any 3DS games at launch because nothing excited me on the 1st party side and I was unfamiliar with nearly all the third party games (not to mention there were so many to choose from at launch). The launch ended up so bad that Nintendo slashed the price within months.

The Wii U...well I got the deluxe bundle and I bought NSMBU...and that and Nintendoland were fine. But that was it really and again, I was completely unfamiliar with all the 16 other 3P games due to no prior experience on GBA, DS, and Wii. And the launch of that system was so bad that Nintendo had that infamous "panic" January 2013 Direct where they laid all there Wii U cards on the table.

And then the Switch exploded out the gate despite being a "Zelda machine" at launch. But that proved to not necessarily be a bad thing. At the end of the day, a single piece of high quality software AND the promise of heavy hitters in the near future and beyond was able to sell the system. Like, yeah it was just Zelda for 2 months, but people knew MK8D was coming soon, Minecraft was after that, Splatoon 2, M+R, "Pokemon" with Pokken, FIFA/NBA, FEWarriors, SMO, XB2. I don't think 2023 will quite be like that for Switch 2, but they'll likely follow a similar strategy (but through enhanced games, not 1P exclusives).

So TotK at launch, Pikmin 4 summer, some HD ports/remasters, DK?, 2D or 3D Mario, a new IP?, DLC, and other various games. I think one of the first big 3P games could be SF6. Then you announce real, modern FIFA, Madden, and everything else to round out the year.

Hogwarts Legacy, Midnight Sons and Genshin Impact would get proper versions. CPR seem keen on the Switch platform, The Witcher 3 and Cyberpunk would get proper ports. I'm sure Activision was eyeing the Switch long before Microsoft chimed in, Call of Duty within a year seems like a no brainer. A proper FIFA 23/24 would help. Elden Ring would sell another few millions, I'm sure they've been talking about it whatever the timeframe is. Every cloud game sold would probably get a nice native version, so all the Resident Evils, Hitman 3, Control, etc, would be a nice boost.

There are a lot possibilities for 3rd-parties.
 
I'll be honest I want this torture to be over. Nintendo can announce the Switch U and I'll accept it....

I want off the roller coaster :(
75ugho.jpg

You really teed that one up.
Not allegedly, it was an official press release, one week before the end of their FY.
I meant the motivation for why they made that press release being alleged to be the press were about to totally give the surprise away.
 
Ok, since we've likely beat the "Zelda-Drake launch" train and all the data that supports or doesn't support it, I thought we might discuss the alternatives for a Drake launch.

  1. Holiday 2023: Launches with a 2D or 3D Mario? Maybe a DK? Luigi's Mansion 4? MPHD? Not sure what else it could be. Recent history has shown Mario to not be the best launch title.
  2. Spring 2024: Probably March, let's be real. I don't see Mario Kart 10 launching a few months after wave 6 of the BCP and no Nintendo console has ever launched with MK10. They'd want a bigger install base ahead of its launch. No Smash (obviously) for several more years at least.
  3. Holiday 2024: MP4? Not sure what big system seller would be left in the Switch's 8th year. Splatoon 4 ain't comin' until 2026/2027 at the earliest. A new IP? For a new generation launch? When was the last time any console manufacturer did that?
  4. Something else, something beyond, nothing at launch. Well the DSi had that in Japan I believe...but even the lite and OLED launched with SOMETHING, even if they weren't "mega" system sellers.
So the more I look and think about it, I can't think of any better title than Zelda to launch the system with. MAYBE Mario, but to wait another 4-5 months after Zelda? When are we gonna get the second, real fruit of this 20+ year Nintendo + Nvidia partnership? 6.5-7 years later and the best they could get out was Mariko? Come on. I feel like I have to force myself to see holiday 2023 launch. Anything later is personally impossible.
 
Great news for Switch 2 is that we’ll get a whole presentation on it! A Pro like system would’ve never guaranteed that. Now, a true successor will be given a presentation and announcement of games for us to all watch.

Also, Nintendo will announce a successor further out compared to a pro like system.
 
Anyway (again), Nintendo has been too weird lately. We know about a title in January, then one in February, then another in May which we know very little about. Nothing between February and May. That's weird from a company the publishes a dozen games every year, specially with the Mario Movie releasing in April. Then there's Pikmin 4 which had a very odd presentation and no gameplay, which (again) is odd for a game releasing in less that one year.

They are either up to something, or they're trolling us big time.

On the software front, Nintendo has announced a game they're publishing in March: Bayonetta Origins: Cereza and the Lost Demon releases March 17, 2023. From my perspective, that only leaves April currently without a scheduled game between now and May, which as you pointed out is when the Mario movie releases.

Personally, I don't think their software release schedule feels weird for 2023 at the present moment. One reason I think this is looking at 2022's first half, Nintendo didn't have a release in February or May, showing they're comfortable with skipping months while doubling up on others. In 2021's first half, Nintendo skipped having a March release; in 2020's, February and April were missed. I could also see more games getting announced in an early 2023 Direct. I may change my mind after the first big Direct of 2023 or once we get through E3, but so far, things feel normal to me. For that reason, I'm not using the announced software lineup to estimate a new hardware release except for thinking that Zelda releases are sometimes tied to new hardware. I understand that my perspective isn't universal, though.

Confirmed:
-January: FE Engage
-February: Kirby's Return to Dream Land Deluxe
-March: Bayonetta Origins
-May: The Legend of Zelda: TOTK
-No month: Pikmin 4
-DLC: FE Engage, MK8DX, XBC3, Splatoon 3 (TBC if the paid expansion is 2023), M+R: Sparks of Hope (iffy because only published by Nintendo in Japan)

Expected based on history:
-Final XBC3 DLC will release as a standalone SKU like Torna
-Pokemon Scarlet/Violet and TOTK will get DLC, possibly Pikmin 4 as well
 
Last edited:
75ugho.jpg

You really teed that one up.

I meant the motivation for why they made that press release being alleged to be the press were about to totally give the surprise away.
Thanks for the laugh :)
I don't post nor follow this thread all that much, my reference to the roller coaster has more to do with the ups and downs of expectations of hardware coming being dashed by nothing happening, or in this case, unclear hearsay about something being cancelled and not being sure which of the rumored hardware we've heard about actually got canned. Feels like i've been on this train since before fami, with the original rumors of a Switch Pro back before new Switch was released.

I've said many times I would rather prefer an announcement, even if the specs aren't what I hoped, so I can just look forward to the games.
 
0
If those images above are true (looks like it)....That's truly disappointing. Totk gonna lag like crazy smh. Not sure what Nintendo is thinking
 
Last edited:
I don't particularly care about discussing new hardware specifics but I'm very worried about the game leaking now.
 
Reminds me of the unveiling of the Dragon Quest XIs bundle which signaled the end of the speculation about an imminent release of a Pro model.
 
0
it's two the upward looking eyes from the Shield we've seen, the closed lotus flower egg thing from the sky area, and two Hylian crest wings, all smashed together into what is likely the Crest of the Sky Kingdom.
 
What's that bug symbol on the left joycon? Zelda enthusiasts get in here right now!
Hidden content is only available for registered users. Sharing it outside of Famiboards is subject to moderation.


Edit:
it's two the upward looking eyes from the Shield we've seen, the closed lotus flower egg thing from the sky area, and two Hylian crest wings, all smashed together into what is likely the Crest of the Sky Kingdom.
Nevermind, there we are 👍
 
0
I think this could lend more credibility to the rumour of Nvidia using Samsung's 3 nm** process node (perhaps for entry level consumer Blackwell GPUs) since TSMC's N3B process node is rumoured to not have good yields.

** → a marketing nomenclature used by all foundry companies
Samsung 5nm for Drake and Samsung 3nm for Drake revision two years after. Believe.


Woah that Zelda theme OLED looks really nice!

Also, I don't think it deconfirms the Drake for a may release with Zelda. We've gotten special editions of consoles (handheld) around the same time frame as their successors successors We'll get a Zelda direct blow out next year, including the OLED model. Likely after fire emblem engage in late January or February. And then sometime later, we may or may not get a Switch 2 announcement. F it doesn't come by end of February, it's likely not gonna come out at May.
 
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