• 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)

maybe a full-on Ocarina of Time remake as a way to get a visual showcase for the next system. shorter development cycle than your normal full-fledged Zelda they could have it out in a few years vs the usual 6/7.
 
maybe a full-on Ocarina of Time remaster as a way to get a visual showcase for the next system. shorter development cycle than your normal full-fledged Zelda they could have it out in a few years vs the usual 6/7.
I always get sad when I hear of ocarina of time/twilight princess/wind waker remasters/remakes to the switch.
The dual screens made the games so seamless and good, being able to swap any item on the fly without opening up the pause screen was amazing. Unfortunately we will miss out on these features with a switch remake/remaster/port
 


On the timestmap, just a quick mention of how "Ultra performance DLSS is going to be needed on the Switch". Nothing major, obviously, but just nice to be remembered that Drake will benefit heavily from AI reconstruction technology.

There is actually a lot of stuff worth picking up here. Recently we were talking about gen-on-gen upgrades to storage. From the video:

Rich: We're not seeing this kind of massive increase in performance plus the price cut we've typically come to expect and I'm struggling to think of other markets where you do still get that...

John: Even in the console space with those extended release periods, the gaps are closing, right? The upgrades gen-on-gen aren't as impressive as they used to, because that's just the state of technology, it's getting harder to innovate in those areas...

Rich: Microsoft sat us down to make a case for the Series S and why they had to do it. The reason why was economic factors... they couldn't cost reduce Series X a few years down the line and make it cheaper, so they couldn't produce what they thought would be an affordable console... there's figures from TSMC showing that the new process node is like 60% more expensive...

The forces that led to the kind of leaps we've grown accustomed to in technology have dried up, and it's an across the board fact. Nintendo, weirdly, is best positioned for a classic next-gen performance upgrade, simply because they're so far behind the curve. But even Nintendo has to deal with these limits just due to also needing to stay a mobile platform. Meanwhile, ancillary components like storage and memory, Nintendo doesn't have the same kind of "catch up" opportunity.

It will be interesting to see how Sony and Microsoft deal with this going forward. Will this generation need to run unusually long to let the hardware space catch up to let a "next gen" console come out at reasonable prices? Or will a more frequent upgrade cycle actually be the correct call, with Playstation and Xbox essentially being "mid range desktop machines" that are constantly refreshed, with Xbox being basically a game pass box, and Playstation being the place to get Sony games two years before they drop on PC?
 
I always get sad when I hear of ocarina of time/twilight princess/wind waker remasters/remakes to the switch.
The dual screens made the games so seamless and good, being able to swap any item on the fly without opening up the pause screen was amazing. Unfortunately we will miss out on these features with a switch remake/remaster/port
likewise when they dropped the Gamepad features from BOTW, looked to be one of the best uses of the Wii U Gamepad in the early demos.
 
at the risk of derailing the thread, the poster in USUM is not a Toxtricity
then what the fuck was it if not a Gen 8 tease!?



all this talk about gimmicks and shit, I think Iwata designed the switch to be flexible as its gimmick. not just in the docked/handheld play, but in how you play with it. VR and add-on controllers and the like. you don't need to attach a gimmick to the tablet itself, it can come as an add-on controller like a joycon
 
maybe a full-on Ocarina of Time remake as a way to get a visual showcase for the next system. shorter development cycle than your normal full-fledged Zelda they could have it out in a few years vs the usual 6/7.
I can see Nintendo use Nvidia tools like RTX Remix and other AI mods to speed development of remakes/remasters. I would love to see a remake of Ocarina of Time.
 
What is this compared to the Switch and PS4?
Switch is 25.6 GB/s (64-bit, regular LPDDR4 at full speed)
PS4's 176 GB/s (256-bit, GDDR5; reverse calculating suggests 5,500 MT/s)
Do note that this sort of thing isn't exactly a 1:1 comparison. The difference in architectures do matter in terms of how performant they can be while having so and so bandwidth.
Wdym by expansion ports? The N64 ram addon?
Famicom's my first guess.
 
then what the fuck was it if not a Gen 8 tease!?



all this talk about gimmicks and shit, I think Iwata designed the switch to be flexible as its gimmick. not just in the docked/handheld play, but in how you play with it. VR and add-on controllers and the like. you don't need to attach a gimmick to the tablet itself, it can come as an add-on controller like a joycon
this is really where I thought they would take it
I expected a bunch of different joycon rail connected peripherals ... but we didn't really get that
 
Zelda being delayed is even a fact, as it was announced for 2022.
My issue is more with the narrative of the next console being delayed. Delaying hardware more than a few months is extremely unlikely, and there was a time, not so long ago, when the majority of this topic used to agree on that.

The only reason the sentiment in the thread has changed I think is because most are trying to reconcile John Linneman's and Nate's comments with T239 being in production / ready for production. I really don't know what would cause Nintendo to outright cancel this hardware, but given how ambiguous it's position was in past rumors, repositioning it might make sense.

It's all still very fuzzy sounding right now (hoping Nate's podcast gives us some answers and positive news). Some of my questions:
  • Just how recent did the 'cancellation' news come through?
  • If developers are being told that it's been cancelled, how would this fit into the 'repositioning' narrative?
    • Wouldn't the messaging be a bit more nuanced than 'cancellation'?
  • What would prompt such a decision in the first place?
    • If it's driven solely by a positioning / branding issue, then I could see a short delay.
    • If it's a recent hardware finding, perhaps it could be longer, but what finding would come through this late?
  • If it is a very long delay, what are the implications of holding off on production of T239 for an entire year?
 
Switch is 25.6 GB/s (64-bit, regular LPDDR4 at full speed)
PS4's 176 GB/s (256-bit, GDDR5; reverse calculating suggests 5,500 MT/s)
Do note that this sort of thing isn't exactly a 1:1 comparison. The difference in architectures do matter in terms of how performant they can be while having so and so bandwidth.

Famicom's my first guess.

Cool, so 102.4 GB/s wouldn't be too bad of a constraint?
 
There is actually a lot of stuff worth picking up here. Recently we were talking about gen-on-gen upgrades to storage. From the video:



The forces that led to the kind of leaps we've grown accustomed to in technology have dried up, and it's an across the board fact. Nintendo, weirdly, is best positioned for a classic next-gen performance upgrade, simply because they're so far behind the curve. But even Nintendo has to deal with these limits just due to also needing to stay a mobile platform. Meanwhile, ancillary components like storage and memory, Nintendo doesn't have the same kind of "catch up" opportunity.

It will be interesting to see how Sony and Microsoft deal with this going forward. Will this generation need to run unusually long to let the hardware space catch up to let a "next gen" console come out at reasonable prices? Or will a more frequent upgrade cycle actually be the correct call, with Playstation and Xbox essentially being "mid range desktop machines" that are constantly refreshed, with Xbox being basically a game pass box, and Playstation being the place to get Sony games two years before they drop on PC?
Such bait for me :p
My bet for what route is sustainable is some form of the latter. Maybe increment 1 node along with the accompanying architectures every ~4 years. Xbox shouldn't have too much trouble swerving that way, brand-wise. Playstation will need a branding reset though.
(I also think that they should probably drop back downward from ~200 watt in a box to... I dunno, somewhere in the 75-125 watt range? Sell themselves as the reasonable/good enough gaming box in a world with what-the-christ-upper-3-digit-watt-halo-PC-setups,)

---

Cool, so 102.4 GB/s wouldn't be too bad of a constraint?
I personally think that it shouldn't be too bad. Would I love 120 or 136.5? Yes. But still, given what we know from the hack, 102.4 seems more workable here, than 25.6 did for the OG Switch.
 
Last edited:
It's all still very fuzzy sounding right now (hoping Nate's podcast gives us some answers and positive news). Some of my questions:
  • Just how recent did the 'cancellation' news come through?
  • If developers are being told that it's been cancelled, how would this fit into the 'repositioning' narrative?
    • Wouldn't the messaging be a bit more nuanced than 'cancellation'?
  • What would prompt such a decision in the first place?
    • If it's driven solely by a positioning / branding issue, then I could see a short delay.
    • If it's a recent hardware finding, perhaps it could be longer, but what finding would come through this late?
  • If it is a very long delay, what are the implications of holding off on production of T239 for an entire year?
the only way this makes sense to me is if NateDrake was extremely wrong in the timing
  • how recent was this cancellation? has to be 2018 when they were planning mariko
  • devs might have been surveyed about an increased clocked mariko, but Nintendo told them they weren't going to go through with it, instead waiting on a new system
  • probably Nintendo seeing the winds from other, future systems and decided they needed to ride those winds to maintain the hardware's position relative to others
  • holding production isn't something one can really do without consequences. you get sent to the back of the line if you get out of place. we saw this happen with auto manufactures, which lead to a massive undershipment of vehicles
 
Gyro has been one of the best 'gimmicks' (if we must use that word). Seamlessly integrates into traditional control schemes, easier for developers to integrate into a game than something like a second screen, works in both docked and portable mode, enables Wiimote / lightgun / fitness / VR experiences, and can be turned off or ignored completely. People have outright demanded third-party Switch ports have gyro. It remains baffling that RE4 Switch does not have it.
Even the PS5 is more bullish on gyro than the PS4 ever was... The only game I remember that used the gyro capabilities of the DS4 was Persona 5 Royal with the darts minigame
 
0
I get what you're saying, but after the success seen with Twilight Princess and infinitely more so with Breath of the Wild, I don't think alignment with new hardware is ever going to be an 'accident' again, nor would I think they'd drop that target all that easily.
There is no way that 2023 was their intended launch year when they announced the game in 2019. When they started dev in 2017, 2020 was the likely target, with 2021 being a reasonable slippage date. The fact that it got to 2023 when new hardware might be announced is, in fact, a total accident.

Nintendo lets the Zelda arrow fly and has no idea where it will land.
 
Wdym by expansion ports? The N64 ram addon?
The Wiimote expansion ports. One of those rare cases where Nintendo put an expansion port on something and it got used. Quite a lot actually - this was the era when literally everyone had a copy of Guitar Hero, and the Wii controllers were cheaper (because half the hardware was in the Wiimote) and better (because they could use the rumble hardware and the minispeaker in the Wiimote for feedback).
 
0
Such bait for me :p
My bet for what route is sustainable is some form of the latter. Maybe increment 1 node along with the accompanying architectures every ~4 years. Xbox shouldn't have too much trouble swerving that way, brand-wise. Playstation will need a branding reset though.
(I also think that they should probably drop back downward from ~200 watt in a box to... I dunno, somewhere in the 75-125 watt range? Sell themselves as the reasonable/good enough gaming box in a world with what-the-christ-upper-3-digit-watt-halo-PC-setups,)
There are some places that dedicated gaming consoles can offer perf wins, like unified memory pools and shared cache designs. And if anyone is teed up to offer bespoke single chip designs at a cost close to mass produced chips, it's AMD.

On the other hand, they're offering subsidized machines right now on the basis that they'll make software sales. Could they offer those same subsidies on a rolling basis, knowing there would be a glut of machines on the second hand market?

It seems likely that the ease of setting up a console, on a design that pops into an entertainment center, with a controller that doesn't suck, and know all the games in the store at least basically work will always have an appeal broader than PC gaming. The question is just for how long consoles can offer a price/performance line that exceeds PCs by any margin.
 
0
There is no way that 2023 was their intended launch year when they announced the game in 2019. When they started dev in 2017, 2020 was the likely target, with 2021 being a reasonable slippage date. The fact that it got to 2023 when new hardware might be announced is, in fact, a total accident.

Nintendo lets the Zelda arrow fly and has no idea where it will land.

We did hear from Zell that he'd heard of 2020 being a target, but he was the only one and I thought others (Nate) questioned if that was ever anything but very early targets. I also think if that game ended up releasing, it would have been vastly different from what's releasing now, and 2023 would see them 2-3 years into development of a new title.

I also never said 2023 was the initial target. I meant to say that with scope creep and other delays, if the timeline even starts to approach landing on new hardware launch dates, I'd assume they start having those serious conversations about it. In that respect I guess you could say it's an 'accident' that its even close to aligning with new hardware. But I'm saying the characterization as an 'accident' just seems too nonchalant, and that if it starts to make sense to launch align the game due to a blown out timeline, that it's probably going to be an important an unwavering decision. They won't mess with their launch title plans.

Or I'm just wrong and as you say they're just totally indifferent and 'letting the arrow fly.'
 
I struggle to believe in holiday 2023 when they could've just delayed zelda another six months
Yeah that's really the only reason I'm holding onto a sliver of hope that it's coming in May. It just feels very odd to not delay Zelda just a few months so it can launch with the console.
 
I still have confidence this thing will be announced before the end of the FY with a release soon after. I simply haven't heard anything to the contrary yet. Zelda contributes to that, I suppose.
 
The only reason the sentiment in the thread has changed I think is because most are trying to reconcile John Linneman's and Nate's comments with T239 being in production / ready for production.
John Linneman didn't say anything about T239 in 2023, or about T239 at all.

I've yet to see someone give a coherent hypothetical timeline that includes a cancelled 2023 upgrade and accounts for the knock-on effect that would have. The assumption just seems to be "hardware in 2024, duh," but I agree with the people who have observed that if something was planned for 2023 and cancelled, that doesn't make 2024 the next believable year (that's just the conclusion of those who always predicted 2024 as the year for new hardware). Let's look at actual possible scenarios.

1. If the hardware was cancelled more than a year ago, it is not the hardware that is/was in development with NVN2 targeting T239, as that hardware and software were still being worked on by Nvidia in February 2022 (and T239 was worked on in some form throughout 2022). If the latter was not cancelled, then we can continue to extrapolate from its development timeline that a 2023 launch is easily possible and even likely.

2. If the hardware was cancelled in the past year (and therefore presumably was the one with NVN2/T239), there would be no time for Nintendo to develop another new kind of hardware to release in 2024. Even if they already had plans for successor hardware, moving those plans up to 2024 would not be possible unless the original plan had been to launch it unrealistically close to the supposed revision, to the point that it would have been in development alongside it,¹ which is not believable for numerous reasons.

So 2024 is really out of the picture in the case of a cancellation, which leaves 2025 or more likely 2026 as the earliest Nintendo could scramble to get some other kind of new hardware out in the honestly quite calamitous scenario of a 2023 upgrade being recently cancelled. Which begs the question: would Nintendo allow this to happen? What exactly could have gone so wrong with the upgraded hardware that it would be better to go at least 4 years² with nothing,³ rather than at worst delaying it to work out the issues?

So if you ask me, there are two possibilities for the talk of refreshes and cancellations.⁴ One, some kind of change of plans happened multiple years ago, leading to the development of NVN2/T239 as we know it, which may explain why early reporting on timing was wrong and why the development timeline seems long if third parties were indeed brought on board in late 2020. Two, a late delay happened in the past year, probably not unlike what happened with the original Switch, which could have moved the release from late 2022 to early 2023, or early 2023 to late 2023. In the absolute worst case of the latter scenario, this is the only thing that could lead to an actual 2024 launch, but that's still not a cancellation (but it is also the kind of thing we would expect to hear about in more than one place).



¹ Not to mention having most of the same features. A "mid-gen refresh" with DLSS and ray tracing followed closely by a "successor" with those same features, with the only notable tech upgrade probably being a Lovelace GPU instead of Ampere? Huh?

² After releasing new models every 2 years since launch, to boot.

³ An OLED Lite counts as nothing.

⁴ Well, there's a third, which is that some kind of true mid-gen refresh was planned for the timeframe of 2019-2021 and was cancelled, but that has pretty much no bearing on what we're discussing now.
 
John Linneman didn't say anything about T239 in 2023, or about T239 at all.

I've yet to see someone give a coherent hypothetical timeline that includes a cancelled 2023 upgrade and accounts for the knock-on effect that would have. The assumption just seems to be "hardware in 2024, duh," but I agree with the people who have observed that if something was planned for 2023 and cancelled, that doesn't make 2024 the next believable year (that's just the conclusion of those who always predicted 2024 as the year for new hardware). Let's look at actual possible scenarios.

1. If the hardware was cancelled more than a year ago, it is not the hardware that is/was in development with NVN2 targeting T239, as that hardware and software were still being worked on by Nvidia in February 2022 (and T239 was worked on in some form throughout 2022). If the latter was not cancelled, then we can continue to extrapolate from its development timeline that a 2023 launch is easily possible and even likely.

2. If the hardware was cancelled in the past year (and therefore presumably was the one with NVN2/T239), there would be no time for Nintendo to develop another new kind of hardware to release in 2024. Even if they already had plans for successor hardware, moving those plans up to 2024 would not be possible unless the original plan had been to launch it unrealistically close to the supposed revision, to the point that it would have been in development alongside it,¹ which is not believable for numerous reasons.

So 2024 is really out of the picture in the case of a cancellation, which leaves 2025 or more likely 2026 as the earliest Nintendo could scramble to get some other kind of new hardware out in the honestly quite calamitous scenario of a 2023 upgrade being recently cancelled. Which begs the question: would Nintendo allow this to happen? What exactly could have gone so wrong with the upgraded hardware that it would be better to go at least 4 years² with nothing,³ rather than at worst delaying it to work out the issues?

So if you ask me, there are two possibilities for the talk of refreshes and cancellations.⁴ One, some kind of change of plans happened multiple years ago, leading to the development of NVN2/T239 as we know it, which may explain why early reporting on timing was wrong and why the development timeline seems long if third parties were indeed brought on board in late 2020. Two, a late delay happened in the past year, probably not unlike what happened with the original Switch, which could have moved the release from late 2022 to early 2023, or early 2023 to late 2023. In the absolute worst case of the latter scenario, this is the only thing that could lead to an actual 2024 launch, but that's still not a cancellation (but it is also the kind of thing we would expect to hear about in more than one place).



¹ Not to mention having most of the same features. A "mid-gen refresh" with DLSS and ray tracing followed closely by a "successor" with those same features, with the only notable tech upgrade probably being a Lovelace GPU instead of Ampere? Huh?

² After releasing new models every 2 years since launch, to boot.

³ An OLED Lite counts as nothing.

⁴ Well, there's a third, which is that some kind of true mid-gen refresh was planned for the timeframe of 2019-2021 and was cancelled, but that has pretty much no bearing on what we're discussing now.


I enjoy your posts LiC. - and you're right, I kind of just said "I guess its 2024"

I just hope you aren't getting too angry at me for missing details or being obtuse. I can feel your seething on having to correct people with each citation and dot point etc. :)
 
We did hear from Zell that he'd heard of 2020 being a target, but he was the only one and I thought others (Nate) questioned if that was ever anything but very early targets. I also think if that game ended up releasing, it would have been vastly different from what's releasing now, and 2023 would see them 2-3 years into development of a new title.

Twilight Princess was announced at E3 2004, given an initial release date of 2005, and delayed to late 2006.
Skyward Sword was announced at E3 2009, given an initial release date of 2010, and delayed until late 2011.
BotW was announced at E3 2014, given an initial release date of 2015, and delayed to late 2016 (and then delayed 3 more months to coincide with the launch of the Switch).
TotK was announced at E3 2019. Then COVID happened.

Given Zell said it was originally intended for 2020, I'm inclined to think precedent was being followed. Had COVID not happened I think TotK would have been Holiday 2021 after originally aiming for 2020.

COVID added another year, so 2022, and then I had thought they pulled a Switch and delayed it to coincide with the Switch 2, but given the OLED leak I'm not so sure.
 
I enjoy your posts LiC. - and you're right, I kind of just said "I guess its 2024"

I just hope you aren't getting too angry at me for missing details or being obtuse. I can feel your seething on having to correct people with each citation and dot point etc. :)
Nah, no way, I'm not going to get angry at anybody's good faith discussion. I just like typing up my thoughts at length, helps me organize them. I'm responding to the broader discussion, too.

If people believe in 2024 for any reason, that's cool, we're all just guessing/predicting. It's just that in the case of the "cancellation" story I see some conclusions that don't quite check out to me. Some plans changed in the past, some hardware was cancelled or delayed, maybe positioning has changed (as you mentioned in the previous post) ⟹ 2024 hardware? Not how I see it, but it's believable. But a specific device, especially the one we've been speculating about all this time, was planned for 2023 and then got cancelled ⟹ 2024 hardware? I dunno about that one.
 
I struggle to believe in holiday 2023 when they could've just delayed zelda another six months

Oh sh*t you're right, it doesn't make any sense to release this thing a couple of months after zelda. Maybe TOTK was never meant for Drake, but the amount of delays put its release too uncomfortably close to Drake's release, which would explain our current situation.

The "no more big games after zelda" would also make sense, just as the lack of any announcement/marketing. 2023 could be another lame duck year a la 2016, just with more support. In any case, we will get confirmation after the financial results release at the end of FY23, around April. Funny enough, the Nintendo NX release date was confirmed in April 2016.
 
2. If the hardware was cancelled in the past year (and therefore presumably was the one with NVN2/T239), there would be no time for Nintendo to develop another new kind of hardware to release in 2024. Even if they already had plans for successor hardware, moving those plans up to 2024 would not be possible unless the original plan had been to launch it unrealistically close to the supposed revision, to the point that it would have been in development alongside it,¹ which is not believable for numerous reasons.

Two, a late delay happened in the past year, probably not unlike what happened with the original Switch, which could have moved the release from late 2022 to early 2023, or early 2023 to late 2023. In the absolute worst case of the latter scenario, this is the only thing that could lead to an actual 2024 launch, but that's still not a cancellation (but it is also the kind of thing we would expect to hear about in more than one place).

Nate's most recent statements seem to imply that the device using Drake as originally intended was "cancelled" in favor of using Drake for another device that is somehow different i.e. the device was cancelled, but the SoC was merely delayed. Given Drake was designed with clear goals in mind, and not as some general purpose chip, I can't imagine how a new device could be significantly different enough to warrant that kind of thing. The only thing I can think of is Nintendo wanting to add some new type of input/sensor/peripheral/gimmick, and change the branding/marketing accordingly.

Is something like that even viable?
 
Nah, no way, I'm not going to get angry at anybody's good faith discussion. I just like typing up my thoughts at length, helps me organize them. I'm responding to the broader discussion, too.

If people believe in 2024 for any reason, that's cool, we're all just guessing/predicting. It's just that in the case of the "cancellation" story I see some conclusions that don't quite check out to me. Some plans changed in the past, some hardware was cancelled or delayed, maybe positioning has changed (as you mentioned in the previous post) ⟹ 2024 hardware? Not how I see it, but it's believable. But a specific device, especially the one we've been speculating about all this time, was planned for 2023 and then got cancelled ⟹ 2024 hardware? I dunno about that one.

Let's throw the idea of anything being cancelled out the window.

From what we 'know', does targeting 2024 for some new hardware using T239 make sense? Or does everything we've heard about T239 basically point to 2023 production.
 
Let's throw the idea of anything being cancelled out the window.

From what we 'know', does targeting 2024 for some new hardware using T239 make sense? Or does everything we've heard about T239 basically point to 2023 production.
Everything we know about T239 suggests it can be in mass production right now and in a product as early as this Spring.

Can, being the operative word. We have no info on when Nintendo wants to start manufacturing or shipping it.
 
Found out that people that works on Calisto Protocol and Horizon Zero Dawn have been recently hired by Retro Studios.


Not sure if relevant but I think Metroid Prime will be a spectacle if given the chance to run on a next gen switch.
We'll see..MP Remaster and MP4 are gonna be Switch games first.. I guess to really take more advantage of switch successor, they might have to limit docked to 720p.
Just imagine, new Nintendo hardware misses May and the BOTW 2 release and you have played through the game already, new Mario is getting a switch release in holiday 2023 and New hardware is announced for March 2024 and the launch title to go with it is Zelda Windwaker UHD. Same as the WiiU port but in 4K as they have nothing else lining up with that time frame.

This is why I think 2023 is still likely.
Yeah the only reason why I can see a holiday Drake Release is because they wouldn't want to kill momentum of the switch too early, and they want to milk it.

So if Drake doesn't release in May with botw 2...
Hopefully Q4 2023.

Drake/Switch 2 gets announced near summer/around E3. New Mario game during the holidays as well as a bunch of third parties. Totk gets a 1440p-4k 60fps treatment and maybe an additional 1080p 30-60fps mode with RT. As a separate game or a $10 upgrade for existing switch owners. And then a bunch of third party games. RDR2, maybe a Cod game, mk8d and Xenoblade 3 dlc.
I would temper my expectations regarding the 4K capabilities of the machine. It should be at best comparable to high-end phones and we're expecting better than a PS5. We don't know how the tensor cores will perform in such a mobile setting, certainly well below their desktop counterparts. Most games will be optimized for handheld, they would need a serious graphical downgrade to suddenly output 4K/60fps.
Youtube, Hulu and other video apps will probably be the only ones able to take full advantage of 4K.
If we're a 5nm node on Switch 2, 3 TFLOPs in docked mode isn't out of the question for the GPU, and that's about as powerful as a PS4 Pro. Drake won't have the bandwidth speed as a PS4 PS4 Pro, but that's okay. As others pointed out, Totk on Drake/Switch 2 is not gonna be a major overhaul/remake treatment. We'll likely get a higher resolution, maybe 60 FPS, and also a good chance of higher draw distance, better lighting and shadows of a switch game. Not a 4k game that is PS4 quality.

The switch itself could have had a 1080p 30fps botw from the GPU, but bandwidth and low CPU clock speeds bottlenecked that from happening. And jumping from a 393 gflops GPU to a more efficient 3 tflops GPU, gives us more than 7.5x GPU speed without counting architecture efficiency improvements in Ampere. So 4k is definitely possible on Switch 2 natively. Maybe 60fps also, or if not--with DLSS should be possible.
 
Chips can arrive in Nintendo's hand as late as March for a May release based on past precedent. That said we would hear much more direct evidence from China before then. if new production is in fact kicking into high gear this month there should be a lot more Chinese rumbling soon.
 
0
Everything we know about T239 suggests it can be in mass production right now and in a product as early as this Spring.

Can, being the operative word. We have no info on when Nintendo wants to start manufacturing or shipping it.

Is there precedent for them to just sit on it in this state, and schedule it's production for a 2024 release?

Just curious how 'certain' some of you all are about 2023 being the year for T239.
 
Let's throw the idea of anything being cancelled out the window.

From what we 'know', does targeting 2024 for some new hardware using T239 make sense? Or does everything we've heard about T239 basically point to 2023 production.
depends on when nintendo even wants to release it. chip development can complete any time before Nintendo wants to put out the system

Is there precedent for them to just sit on it in this state, and schedule it's production for a 2024 release?

Just curious how 'certain' some of you all are about 2023 being the year for T239.
the kicker is that this is a semi-custom chip. such a spend would mean Nintendo probably wants development to line up with production of the system since it's their money they're spending on making the chip and they would want returns sooner than later
 
I'm not giving up on May 2023 until one of the following:
  • we see another device using T239 (might be hard to ascertain)
  • we get to February with no leaks
I'm giving it until mid February at the latest myself.. Otherwise I think it's pushed to Q4 2023.

if switch 2 is pushed to fall 2023,Nintendo must be thinking
1. They want to maximize switch console sales as possible with Totk
2. They don't think Totk is needed and would rather go with Mario or another game... But they will likely get us a better version of Totk launch
3. They're not ready yet and would also want more Switch 2 consoles shipped, as well as a bigger launch of games?
OoT 3D getting the Portal treatment seems ripe for a disaster considering how badly Portal RTX runs on a 2080 though, lol.
That game isn't optimized well. That shouldn't prevent a switch game from RTX with decent performance either.

Cool, so 102.4 GB/s wouldn't be too bad of a constraint?
If we compare it to the ps4 and getting 1 to 1 PS4 ports, we should be fine. Nvidia is really good at optimizing with less bandwidth vs AMD GPUs. Especially Ampere cards. We're also getting cache, which should be more than PS4, which should help out a lot. Finally, DLSS should also help too.

The more bandwidth the merrier though. Even 33% more from lpddr5x would go a long way to making it more future proof.

I think that from top to bottom (LPDDR5-6400 to LPDDR5X-8500) is from most likely to least likely to be used for Drake.
Yeah, we'll likely barely miss it, even if we get it in Q4 2023,for cost savings purposes. Nintendo probably figures DLSS should close the gap in some areas as well.

Lpddr5x and a more efficient node almost sound like a given for a revision with same clock speeds as lpddr5 to reduce power draw...
 
Oh sh*t you're right, it doesn't make any sense to release this thing a couple of months after zelda. Maybe TOTK was never meant for Drake, but the amount of delays put its release too uncomfortably close to Drake's release, which would explain our current situation.

The "no more big games after zelda" would also make sense, just as the lack of any announcement/marketing. 2023 could be another lame duck year a la 2016, just with more support. In any case, we will get confirmation after the financial results release at the end of FY23, around April. Funny enough, the Nintendo NX release date was confirmed in April 2016.
I really hope Nintendo does the same thing this year if nothing is announced by then. If all Nintendo says is that the next generation of Switch is coming March 2024 and nothing else, i can be completely content with that, since i will know it is real and when to expect it and save up my money
 
Yeah, we'll likely barely miss it, even if we get it in Q4 2023,for cost savings purposes. Nintendo probably figures DLSS should close the gap in some areas as well.

Lpddr5x and a more efficient node almost sound like a given for a revision with same clock speeds as lpddr5 to reduce power draw...
I think timing rather than costs is probably the reason Nintendo uses LPDDR5 instead of LPDDR5X.
Samsung's LPDDR5X-7500 modules were announced to be validated by Qualcomm on 3 March 2022.
And if the LinkedIn profile of a hardware engineer working at Nvidia is any indication, T239 was probably taped out sometime between early 2022 and mid 2022.
That's important since the choice of what memory controller is used for Drake is finalised when Drake's taped out since the memory controller is generally integrated into the SoC.
And there's a possibility the memory controller for LPDDR5 is not forwards compatible with LPDDR5X, similar to how the memory controller for LPDDR4 is not forwards compatible with LPDDR4X.

But I do agree that LPDDR5X being used for hardware equipped with a die shrink of Drake is definitely a possibility.
 
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