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

[Goodtwin]

I think the increase in production is actually suspicious.

Nintendo is running out of people to sell the Switch to. I would expect 20% decrease in hardware forecast, not a 10% increase. The whole industry is down this year. I don't think a console on its 7th year would sell 20+ millions, with or without a 3D Mario, with or without a price cut.

I think it would take a significant price-cut, a huge push in peripheric markets, a 3D Mario and a game that goes viral, such as Nintendogs, for Nintendo to sell that many Switches.

That increased forecast hides something else.

We have to be approaching that soon right? There are always new gamers entering the market, but that only accounts for maybe 25% of their year to year sales. Eventually every teenager and adult who wanted a Switch has already bought one. Yes the OLED may get some people to double dip and upgrade but I cant imagine the percentages are super high for that scenario. More likely than not, Switch sales will be down year over year. Software sales could remain strong thanks to good sales for titles like Zelda TotK and Pikmin 4, but even for games that those we may see sales lower than expected even though the userbase for Switch is massive. Eventually you start to lose some of those early Switch adopters who have moved on to newer devices. If there is a new 3D Mario game for the holiday season later 2023 along with a price cut, maybe Nintendo can move 20 million Switch units, but that would be very optimistic and the likelihood of sales dropping significantly in 2024 go way up since Nintendo will have burnt through some of their heaviest hitters in 2023.

Nintendo could very well be gearing up to stock pile Switch units so that they can transition those manufacturing partners to start creating their next hardware. There is a finite level of capacity for these manufactures. I believe this is part of the reason why Sony halted PS4 production pretty early on so they could have their manufacturing partners focused on PS5 units.

 

[Dekuman]

Is 256 GB of storage more likely now?

not an expert, but i think it comes down to their approach to digital. if the goal is to have games load off microSD perhaps tapping into the higher read speeds currently not used by Switch, then they may stick to 64GB to save costs.

i think this is the most likely scenario, the alternatives are simply too expensive or exotic for a mass market product.
They may simply enforce a stricter guidelines of high speed SDs being a requirement. Currently. the 10mb/s transfer speed isn't a big ask.

 

[ILikeFeet]

Is 256 GB of storage more likely now?

yes. though will depend on how much Nintendo wants to spend. though I think this will cost more than the 64GB Switch OLED, so I think we will get 128GB at minimum

 

[Teal'c]

Ram is one of the things they can easily change last minute. A lot of devs thought the PS4 would have 4gb until the reveal.

I don't know what else they could change.

As someone pointed out, they might change the amount of internal memory depending on the cost.
Maybe not the type for several factors: form factor, TDP etc.

it could be that production wasn't scheduled to start yet in the first place. the chip being done is kinda independent of that.

node changing might have been something decided on a year or more ago. delaying a device to a year doesn't give you a lot of time to make changes. a year is more in the "the headlining game isn't ready yet" territory. more ram is doable at least

So if the delay is only one year (assuming March 2024) the change of node is ruled out... assuming March 2025 instead?

I wonder because if the 'cancellation' was confused with a node change (8nm to something else) maybe it has already happened.

 

[arsen]

The 20M Switch are for the 20M peoples who played BOTW on Cemu who will want to play it on a real Switch because of how dogshit Switch emulators are compared to Cemu

 

[Mr Swine]

I still have a hard time believing that Nintendo/Nvidia would waste several hundred million on R&D for something and then go “well we have Drake” and have 2 different SoC so close together?

Either Nintendo is ditching Switch for something completely new, no BC and releasing in 2024 or they have just decided to milk Switch for all it’s worth with no price cut and wait for Switch 2

 

[Hermii]

As someone pointed out, they might change the amount of internal memory depending on the cost.
Maybe not the type for several factors: form factor, TDP etc.

So if the delay is only one year (assuming March 2024) the change of node is ruled out... assuming March 2025 instead?

I wonder because if the 'cancellation' was confused with a node change (8nm to something else) maybe it has already happened.

It's unrealistic they would ever consider 12sm on 8nm imo, so if there was a node change it was prior to the Nvidia leak.

 

[ILikeFeet]

So if the delay is only one year (assuming March 2024) the change of node is ruled out... assuming March 2025 instead?

I wonder because if the 'cancellation' was confused with a node change (8nm to something else) maybe it has already happened.

the "cancellation" can't be confused with nodes because the news came from devs. they wouldn't know the node because it's irrelevant to them.

 

[10k]

There is no cancelation. There's a pivot. Copium

 

[Serif]

Damn, the report about Nintendo asking developers to make their games "4K-ready" was all the way back in Sept. 2020.

Still wondering if Mochizuki's heard anything about a postponed device, or if he's just waiting for more sources to corroborate. Taking a while, if that postponement happened mid-Summer / Fall 2022.

 

[pelusilla6]

Damn, the report about Nintendo asking developers to make their games "4K-ready" was all the way back in Sept. 2020.

Still wondering if Mochizuki's heard anything about a postponed device, or if he's just waiting for more sources to corroborate. Taking a while, if that postponement happened mid-Summer / Fall 2022.

He answered with a "" to John's podcast about something cancelled.

And last year he tweeted the naughty "2 early"

 

[Eclissi]

I believe that from the next Direct we will understand something.
Nintendo can't think of supporting Switch for 1.5 years (after Zelda) with only Pikmin 4 and maybe MP4, and then a series of DLC and remaster... This would lead to an inevitable loss of momentum.
If there are no heavy announcements in the Direct, then Drake is still 2023 for me.

 

[Raccoon]

I believe that from the next Direct we will understand something.
Nintendo can't think of supporting Switch for 1.5 years (after Zelda) with only Pikmin 4 and maybe MP4, and then a series of DLC and remaster... This would lead to an inevitable loss of momentum.
If there are no heavy announcements in the Direct, then Drake is still 2023 for me.

ehhh quiet twilights are nothing new for nintendo

throw 2D Mario in with some small stuff and you've got 2023 and 2024 covered

 

[10k]

I believe that from the next Direct we will understand something.
Nintendo can't think of supporting Switch for 1.5 years (after Zelda) with only Pikmin 4 and maybe MP4, and then a series of DLC and remaster... This would lead to an inevitable loss of momentum.
If there are no heavy announcements in the Direct, then Drake is still 2023 for me.

I’m sure they got a few remasters planned, another Kirby project, Captain Toad 2 and a DK game coming. Maybe the next Mario is also cross gen.

 

[kalosmani]

The 20M Switch are for the 20M peoples who played BOTW on Cemu who will want to play it on a real Switch because of how dogshit Switch emulators are compared to Cemu

Zelda release will probably improve Switch emulator efforts just as how Breath of the Wild made the dedication to get a decent Wii U emulator going.

 

[ShaunSwitch]

Anybody got any backup plans for their switch 2 budget for this year?

I don't want to get an OLED model because I generally find handheld play quite uncomfortable and use docked and table top more, but wouldnt mind a portable OLED experience, don't want to buy an OLED if a switch 2 is coming in a year as well.

So I am thinking of grabbing one of those portable 4k AMOLED displays from china that come with a built in switch Dock, I figure it will be great for a drake switch and also upgrade my current switch experience, plus I can use it with my phone for retro emulation and as a second laptop display.

 

[10k]

Anybody got any backup plans for their switch 2 budget for this year?

I don't want to get an OLED model because I generally find handheld play quite uncomfortable and use docked and table top more, but wouldnt mind a portable OLED experience, don't want to buy an OLED if a switch 2 is coming in a year as well.

So I am thinking of grabbing one of those portable 4k AMOLED displays from china that come with a built in switch Dock, I figure it will be great for a drake switch and also upgrade my current switch experience, plus I can use it with my phone for retro emulation and as a second laptop display.

PS5 DE most likely.

 

[Concernt]

Is 256 GB of storage more likely now?

Why would it be?

I don't think it was ever extremely unlikely, but what about "now" has implied it may be more likely?

 

[Concernt]

Anybody got any backup plans for their switch 2 budget for this year?

I don't want to get an OLED model because I generally find handheld play quite uncomfortable and use docked and table top more, but wouldnt mind a portable OLED experience, don't want to buy an OLED if a switch 2 is coming in a year as well.

So I am thinking of grabbing one of those portable 4k AMOLED displays from china that come with a built in switch Dock, I figure it will be great for a drake switch and also upgrade my current switch experience, plus I can use it with my phone for retro emulation and as a second laptop display.

Save it for Switch 2 the next year, orrrr go on holiday.

As for your plans: absolutely do not get a "4K AMOLED display with build in dock for NSwitch" with the hope it'll work with Drake. Third party docks from 2020 didn't necessarily support OLED Model.

 

[Cybergatuno]

Anybody got any backup plans for their switch 2 budget for this year?

I don't want to get an OLED model because I generally find handheld play quite uncomfortable and use docked and table top more, but wouldnt mind a portable OLED experience, don't want to buy an OLED if a switch 2 is coming in a year as well.

So I am thinking of grabbing one of those portable 4k AMOLED displays from china that come with a built in switch Dock, I figure it will be great for a drake switch and also upgrade my current switch experience, plus I can use it with my phone for retro emulation and as a second laptop display.

I got a V1 Switch that I barely touch nowadays. The only game on Switch that I look forward to is TotK (and The Plucky Squire). I love playing handheld while watching a game or a movie with the family, but that game is getting my full attention on TV. So I don't have much of a reason to upgrade. Game Pass has been filling my empty time otherwise.

The only thing that would get me to buy an OLED (and NSO) is if TotK has online story co-op so I would play with family.

 

[ShaunSwitch]

Save it for Switch 2 the next year, orrrr go on holiday.

As for your plans: absolutely do not get a "4K AMOLED display with build in dock for NSwitch" with the hope it'll work with Drake. Third party docks from 2020 didn't necessarily support OLED Model.

It's got HDMI too, so plan is to get a secondary drake Dock when drake does launch to use with it.

 

[Concernt]

It's got HDMI too, so plan is to get a secondary drake Dock when drake does launch to use with it.

Oh fair. Good luck!

 

[Deleted member 887]

It's unrealistic they would ever consider 12sm on 8nm imo, so if there was a node change it was prior to the Nvidia leak.

The power draw equation kinda doesn't add up for me at all, no matter what.

Lovelace's documented power draw is something like 6W/TFLOP. Lovelace is not only on N4, but it has additional power saving tech above what's in Ampere, and yet it seems to be just about par on the power front.

My suspicion is that Lovelace's power draw wins from clock gating (which seem to have been back ported to Drake) are getting sucked up by the double rate OFA, if not also the updated Tensor and RT cores. I think DLSS 3 is out of the question for a device as small as Drake, even with that updated OFA, and Lovelace didn't update the raster pipeline in any real way over Ampere.

If so, this might answer the question of "if Drake is 5nm why isn't it Lovelace?" They took the things they could use from Lovelace - the NVENC and clock gating updated - left the rest behind. I am dubious that those things can offer decent power draw on 8nm by themselves, but would explain why Drake would stay on Ampere even if moved elsewhere.

 

[NineTailSage]

Who does a 20 million final year?

Doesn't have to be a new user-end gimmick, but ~6 years of change in CPU/GPU/RAM/etc. is something slapping an updated SDK onto old hardware isn't going to properly convey. The changes go a lot beyond: Just develop a Switch game, then make it high res with an updated kit.

I probably should have clarified but when I stated SDK I was meaning it as Software Development Kit, since most people's assumptions are that games created for this new device will also probably run on the current Switch for the 1st year or two. I still stand by hardware companies not needing to ship out every variation of silicon like they use to generations ago, and can work on both hardware and software in tangent without one disrupting the other...

 

[Foltzie]

So here's the biggest thing I don't understand about the whole dev-kits fiasco and I will preface this by saying I'm not a developer so things could be just over my head. The earlier rumors of the NX specs were thought to be higher end hardware because the NX tech demos needed cutting-edge chips to run them. Which we then found out later that the Nvidia TX1 was chosen as the final hardware and it wasn't what those first rumors suggested.

Im not an expert, I'll leave that to the Thraktor's and Old Puck's, but I would presume the original Switch dev-kits would have been Windows boxes with Maxwell based GeoForce cards using the early version of the NVN library. At the very least the hardware would have to be sized up to account for Windows overhead. For example they may have had to have a 900 series card in order to match the memory bandwidth expected in the TX1.

I get that earlier console dev-kits from manufacturers were sent out in multiple waves, but with the advancements in development environments and software, most of this process has changed quite a bit from the 5th, 6th and 7th generation days. The Switch is an established gaming platform and has a understandable development environment, so why would Nintendo and Nvidia realistically need to send out early build dev-kits if the main changes from Switch to Switch 2 is 4k(upscaled) graphics. It seems like the same results can be achieved with a newer SDK using either Turing or Ampere level cards until more finalized hardware rolls off the line. Again unless the Switch 2 has some wild new gimmick that needed to be expressed in actuality, we fully understand how the switching method is achieved...

Both Ampere cards and Orin are using SEC's 8nm node and it seems between the two of these lines of products Nvidia and Nintendo can get a good simulation environment of what that process could achieve, before actually spending hundreds of millions of dollars in R&D.(designing a custom SoC that they would scrap or need to modify to a different node)

I dont know if ongoing Windows based NVN development today requires a machine provided by Nvidia/Nintendo or if there are supported configurations posted on the developer site. However, I would presume new machines supporting NVN2 would definitely be provided by Nvdia/Nintendo to reduce variables during development of the final hardware and supporting software.

 

[ITSMILNER]

Nintendo can easily coast to the end of 2024 with TotK, MP4, Pikmin 4, some HD Remasters and DLC

 

[Teal'c]

the "cancellation" can't be confused with nodes because the news came from devs. they wouldn't know the node because it's irrelevant to them.

I see. I thought they confused (or even misinterpreted) the withdrawal of the devkits for something bad, but they don't actually care about the node on which the chip is manufactured.

 

[10k]

Nintendo can easily coast to the end of 2024 with TotK, MP4, Pikmin 4, some HD Remasters and DLC

That's not going to sell 21m Switch units.

 

[Hermii]

The power draw equation kinda doesn't add up for me at all, no matter what.

Lovelace's documented power draw is something like 6W/TFLOP. Lovelace is not only on N4, but it has additional power saving tech above what's in Ampere, and yet it seems to be just about par on the power front.

My suspicion is that Lovelace's power draw wins from clock gating (which seem to have been back ported to Drake) are getting sucked up by the double rate OFA, if not also the updated Tensor and RT cores. I think DLSS 3 is out of the question for a device as small as Drake, even with that updated OFA, and Lovelace didn't update the raster pipeline in any real way over Ampere.

If so, this might answer the question of "if Drake is 5nm why isn't it Lovelace?" They took the things they could use from Lovelace - the NVENC and clock gating updated - left the rest behind. I am dubious that those things can offer decent power draw on 8nm by themselves, but would explain why Drake would stay on Ampere even if moved elsewhere.

Not pretending to be an expert, but desktop cards really aren't clocked for power efficiency though, they are clocked for performance. So not sure they are the best benchmark for Drake.

 

[YugamiSekai]

Anybody got any backup plans for their switch 2 budget for this year?

I don't want to get an OLED model because I generally find handheld play quite uncomfortable and use docked and table top more, but wouldnt mind a portable OLED experience, don't want to buy an OLED if a switch 2 is coming in a year as well.

So I am thinking of grabbing one of those portable 4k AMOLED displays from china that come with a built in switch Dock, I figure it will be great for a drake switch and also upgrade my current switch experience, plus I can use it with my phone for retro emulation and as a second laptop display.

I think I'm just gonna buy the OLED tablet itself from somewhere, I don't really need the dock or the joycons, but I have to get some kind of Switch this year or else my save data backups are going in the dumps.

 

[ILikeFeet]

I see. I thought they confused (or even misinterpreted) the withdrawal of the devkits for something bad, but they don't actually care about the node on which the chip is manufactured.

I would find it weird if the (very few) devs that are allowed access aren't given a heads up on the happenings of dev kits. having them (or sdks or whatever) taken and not being told if they're being updated or something would cause some worse-case-scenario assumption.

 

[karmitt]

Nintendo can easily coast to the end of 2024 with TotK, MP4, Pikmin 4, some HD Remasters and DLC

They can - as in they’ll survive it. But coasting on years 7-8 does not sound like a great way to keep momentum.

I just cannot see that being their plan. Pikmin and Metroid could be exceptional entries and I still doubt Nintendo would bank on them keeping consumer interest high. DLC certainly isn’t going to move the needle. And jury’s out on these “HD Remasters” - we’ve yet to see any of them or the efforts put into them. My gut says anything but remakes won’t matter all that much, especially for the titles being discussed (Baten Kaitos, FZero). And at some point Nintendo will need to line up major products to be ready for launch, cross-gen or otherwise.

I’m genuinely wondering if the increase in sales is lumping new and old hardware together, calling it all Switch. 21 million feels like a reasonable number if we see Switch 2 launch later this year (not going to speculate FY or CY). Switch (1) sales will continue to fall, but marketing alongside Switch 2 with a price cut would keep it somewhat relevant.

 

[mjayer]

Nintendo can easily coast to the end of 2024 with TotK, MP4, Pikmin 4, some HD Remasters and DLC

My guess it that Pikmin 4 and MP4, and the next big Mario game will be held back for the launch of Switch 2, assuming it launches in 24. Perhaps not Pikmin 4 but I would be surprised. And that F-Zero rumor? Gotta be targeting the Drake with a down port incoming as well. It just seems like they are holding back so much right now...

I'm sill currently 75% positive we'll get Switch 2 in 23 though.

 

[Dekuman]

They can - as in they’ll survive it. But coasting on years 7-8 does not sound like a great way to keep momentum.

I just cannot see that being their plan. Pikmin and Metroid could be exceptional entries and I doubt Nintendo would bank on them keeping consumer interest high. DLC certainly isn’t going to move the needle. And jury’s out on these “HD Remasters” - we’ve yet to see any of them or the efforts put into them. My gut says anything but remakes won’t matter all that much, especially for the titles being discussed (Baten Kaitos, FZero). And at some point Nintendo will need to line up major products to be ready for launch, cross-gen or otherwise.

I’m genuinely wondering if the increase in sales is lumping new and old hardware together, calling it all Switch. 21 million feels like a reasonable number if we see Switch 2 launch later this year (not going to speculate FY or CY). Switch (1) sales will continue to fall, but marketing alongside Switch 2 with a price cut would keep it somewhat relevant.

I think they have big plans for Switch this year, with our without new hardware.

Their stated goal of smoothing out earnings and avoiding peaks and valleys in earnings would weigh heavily. I don't think they are coasting because they are trying to maximize profit, if the revision was canned for a full successor, it sounds like they are trying to deliver the best product possible. it just sucks to wait this long. At least with other platforms or even previous Nintendo handhelds, power users got a 'pro' model to tide them over until the succ launched.

 

[karmitt]

I think they have big plans for Switch this year, with our without new hardware.

Their stated goal of smoothing out earnings and avoiding peaks and valleys in earnings would weigh heavily. I don't think they are coasting because they are trying to maximize profit, if the revision was canned for a full successor, it sounds like they are trying to deliver the best product possible. it just sucks to wait this long. At least with other platforms or even previous Nintendo handhelds, power users got a 'pro' model to tide them over until the succ launched.

I agree that there has to be big plans, but I personally cannot fathom what plans they could have that would see interest rise dramatically, fighting typical trends. Price cuts, unless steep(?), don’t seem like enough, but I could very well be wrong on that one.

Then what software could they possibly have in store that’s going to see Nintendo have an unprecedented year? And why would they be keen to launch software that has that much of an impact so close to new hardware? They’re already skipping over Zelda; Just how many heavy hitters will they relegate to this generation?

I do hope we get some answers from somewhere soon. The next Direct should be at least somewhat telling.

 

[Disorientator]

Has anyone taken a closer look at the NERD AI-upscale patent application that was published last week? (filed July 13th, 2021)

I think it was brought-up by someone at one point, but not sure if it got any traction.

It seems to be based on the older NERD patent applications (filed March 25th, 2020) already discussed, and as such both of the original patents "are incorporated by reference", but upon closer inspection there are some slight variations to the original text/descriptions that caught my eye.

Original (2020)

In certain examples, multiple neural networks may be provided with the same instance of a game (e.g., an individual download or specific physical media instance) to allow for the game to output to different types of displays (e.g., 1080p in one instance, 1440p in another, 4k in another, etc.)

New (2021)

In certain examples, multiple neural networks may be provided for the same instance of a game (e.g., an individual download or specific physical media instance) to allow for the game to output to different types of displays (e.g., 1080p in one instance, 1440p in another, 4k in another, etc.) and/or under different processing conditions. For example, a powerful computer (e.g., with more powerful hardware components) may acquire a neural network that allows for game images rendered at 1080p to be upscaled to 4k. In another example, a less power computer (e.g., with less powerful hardware components than the above example) may acquire a neural network that allows for game images rendered at 540p to be upscaled to 1080p.

As you can see the "new" patent application goes a step further and, even as just an example, specifically mentions "less" and "more" powerful systems.

Could they be talking about less and more powerful systems in the Switch or Switch-Next families (e.g. Switch Pro or 2 - this is NERD after all), or is it just a generalization about any less and more powerful systems, being a patent and all that?

Also, is the "a powerful computer may acquire a neural network that allows for game images rendered at 1080p to be upscaled to 4k" part talking about game patches?

 

[Goodtwin]

21 million feels like a reasonable number if we see Switch 2 launch later this year (not going to speculate FY or CY). Switch (1) sales will continue to fall, but marketing alongside Switch 2 with a price cut would keep it somewhat relevant.

We will have to wait till April to find out what Nintendo estimates for the next fiscal year. The article about suppliers gearing up for increased production may be true even if Nintendo doesn't plan to sell that many units. Nintendo spent a ton of money stocking up on raw materials this fiscal year compared to the last despite selling fewer units.

Then what software could they possibly have in store that’s going to see Nintendo have an unprecedented year? And why would they be keen to launch software that has that much of an impact so close to new hardware? They’re already skipping over Zelda; Just how many heavy hitters will they relegate to this generation?

Nintendo would empty the cupboard this year with Zelda TotK, a 2D and 3D Mario game, Metroid Prime Trilogy HD and Metroid Prime 4. They absolutely could have the best year of software releases ever, but then what? At some point you do need to have software ready to push your next hardware. Maybe some of these titles releasing have actually been completed for some time now and they are waiting to strategically release them and have already moved on to developing new games for the next Switch? It will be interesting to see what comes from the next Nintendo Direct. What they announce should help paint a clearer picture of what their roadmap really looks like.

 

[Sol]

May 2023: Tears of the Kingdom
September 2023: Pikmin 4
October - December 2023: Pokemon Scarlet Violet expansion, new Mario Party, Metroid Prime remake
Spring 2024: New 3D Donkey Kong
Summer 2024: New Star Fox
Fall 2024: Metroid Prime 4, new Pokemon game remake (Gen 5/Let's Go Johto)

Sprinkle in some Gamecube remakes, Zelda remakes, and other titles like Mario sports games, 2D Mario, etc, and the Switch can easily make it to Spring 2025. The wait would suck, but Nintendo could absolutely do it without any issues.

 

[YolkFolk]

My guess it that Pikmin 4 and MP4, and the next big Mario game will be held back for the launch of Switch 2, assuming it launches in 24. Perhaps not Pikmin 4 but I would be surprised. And that F-Zero rumor? Gotta be targeting the Drake with a down port incoming as well. It just seems like they are holding back so much right now...

I'm sill currently 75% positive we'll get Switch 2 in 23 though.

Wasn’t the F-Zero rumour a remaster of F-Zero GX?

 

[ILikeFeet]

Has anyone taken a closer look at the NERD AI-upscale patent application that was published last week? (filed July 13th, 2021)

I think it was brought-up by someone at one point, but not sure if it got any traction.

It seems to be based on the older NERD patent applications (filed March 25th, 2020) already discussed, and as such both of the original patents "are incorporated by reference", but upon closer inspection there are some slight variations to the original text/descriptions that caught my eye.

Original (2020)


New (2021)


As you can see the "new" patent application goes a step further and, even as just an example, specifically mentions "less" and "more" powerful systems.

Could they be talking about less and more powerful systems in the Switch or Switch-Next families (e.g. Switch Pro or 2 - this is NERD after all), or is it just a generalization about any less and more powerful systems, being a patent and all that?

Also, is the "a powerful computer may acquire a neural network that allows for game images rendered at 1080p to be upscaled to 4k" part talking about game patches?

for the sake of the patent, they're probably generalizing as much as possible. as for "acquire a neural network" it could mean anything. from patches to integration during development

 

[Raccoon]

for the sake of the patent, they're probably generalizing as much as possible.

paging @Skittzo to explain why the system is broken

 

[Skittzo]

paging @Skittzo to explain why the system is broken

Ilikefeet did a pretty good job at that. Companies will often try to broaden the coverage of their patents once they see one work.

 

[Deleted member 887]

Also, is the "a powerful computer may acquire a neural network that allows for game images rendered at 1080p to be upscaled to 4k" part talking about game patches?

The patents are the same, they've extended the use cases for the patent to cover more stuff, but the base claims are identical. This particular trick could be done via patches, or by shipping multiple neural networks in the original game.

The idea is that you have the same original resolution on every model of hardware, but more powerful hardware can run it at a faster framerate. That leave more time to run a deeper CNN, and do higher quality upscaling.

This is a fairly obvious extension of the patent's technology, but often patents list out as many applications as possible. This isn't about making the patent broad, but keeping someone from sniping a only-tiny-difference patent on the same tech later.

 

[Nyrnvs]

Damn, the report about Nintendo asking developers to make their games "4K-ready" was all the way back in Sept. 2020.

Still wondering if Mochizuki's heard anything about a postponed device, or if he's just waiting for more sources to corroborate. Taking a while, if that postponement happened mid-Summer / Fall 2022.

Mochizuki being oddly quiet on all this right now is really weird to me. One of the top Nintendo journalists having simply posted an emoji to DF’s podcast on Twitter is…head scratching.

Seems this would be big news for Bloomberg or other publications he’s had bylines on…yet nothing. I don’t know if he’s just very picky on what he covers following the OLED/Zynga debacle or everyone’s got their cards extremely close to their vest right now.

 

[Disorientator]

for the sake of the patent, they're probably generalizing as much as possible. as for "acquire a neural network" it could mean anything. from patches to integration during development

Companies will often try to broaden the coverage of their patents once they see one work.

The patents are the same, they've extended the use cases for the patent to cover more stuff, but the base claims are identical. This particular trick could be done via patches, or by shipping multiple neural networks in the original game.

The idea is that you have the same original resolution on every model of hardware, but more powerful hardware can run it at a faster framerate. That leave more time to run a deeper CNN, and do higher quality upscaling.

This is a fairly obvious extension of the patent's technology, but often patents list out as many applications as possible. This isn't about making the patent broad, but keeping someone from sniping a only-tiny-difference patent on the same tech later.

Excellent input as usual, cheers guys.

 

[AlwaysOnDeck]

Nintendo can easily coast to the end of 2024 with TotK, MP4, Pikmin 4, some HD Remasters and DLC

Especially if they finally do a price drop.

 

[AlwaysOnDeck]

I don't want to get an OLED model because I generally find handheld play quite uncomfortable and use docked and table top more.

The OLED is made more, for tabletop, than handheld.

 

[fwd-bwd]

I saw that the latest Bloomberg article was misquoted a few times—Mochizuki and Wu did not write that Nintendo plans to ship 21 million Switches in FY03/2024. Here are the key points that they reported:

• Although Nintendo forecasted to ship 19MM units in FY03/2023, multiple sources indicated that the actual shipment will be roughly 21MM.
• Suppliers and assemblers were told to expect making "more Switch units" in FY24 than FY23, but the exact number wasn't communicated.
• The report did not specify whether the FY24 forecast is "more" than 19MM (FY23 forecast) or 21MM (actual shipments).
• You may recall the Nikkei opinion piece from last week, in which the author claimed that Nintendo's supply chain orders suggest 20MM units for FY24. (>19MM)
• On the other hand, if the Bloomberg report actually meant >21MM, the target could be 22MM units or more.

Be it 20MM or 22MM, I find the number stupendous for a console entering its 7th year, and I aren't the only one. Some have theorized that Nintendo might reduce the prices of all models by $50 to stimulate the sales. That does not seem a winning strategy to me though.

Spoiler: Profit/loss from price cut

• Let's assume that Nintendo ends up selling 21MM units (the median between 20MM and 22MM) after a $50 cut of the retail price.
• The retailer markup for a console is usually 5%-10% (yes, that's low, and the reason why they want you to buy accessories and stuff). Let's say 7.5%, and that translates to a $46.25 reduction of net profit per console for Nintendo.
• To calculate the profit/loss resulting from a price cut:
  • N = number of units sold if no price cut
  • 21MM - N = additional units sold if price cut
  • P = avg net profit per unit if no price cut
  • P - $46.25 = avg net profit per unit if price cut [IRL the ratio of OLED, v2, and Lite sales may change after a price cut, thus altering P; but for simplicity I keep it constant here]
  • N*P = total net profit if no price cut
  • 21MM*(P - $46.25) = total net profit if price cut
  • D = 21MM*(P - $46.25) - N*P = incremental profit/loss due to price cut

As you can see from the middle of Table 1, when N=16MM and P=$90, a $50 retail price cut would generate 5MM additional Switch sales, but reduce Nintendo's net profit to the tune of half a billion dollars (-$521MM)! To break even (D=0; see Table 2), when P=$90, N needs to be about 10.21MM and the price cut must spur 10.79MM additional sales.

What if, you ask, the easing of supply chain constrains lowers the production and shipping costs in FY24? Even when let's say the unit cost magically decreases by $30 from FY23 to FY24 (P thus rising from $90 to $120), it'd still be a loss unless N≤12.91MM and incremental sales >8.09MM.

The rest of Tables 1 and 2 demonstrate some other N and P combinations, and the numbers look dire. Clearly, reducing the retail price by $50 is not a good business idea, unless the pre-price-cut profit is sufficiently high and the stimulated incremental sales are unrealistically large.

Of course Nintendo might be willing(?) to take a bath on hardware profits, and earn it back via software sales. However, I'm not too optimistic regarding the LTV of price-sensitive late adopters. And when a hardware unit is purchased by a household with 1+ existing Switches, its tie ratio may be close to zero.

Lastly, the Switch hardware market essentially is a durable goods monopoly. It does not benefit the monopolist (Nintendo) to price cut, because that'd be a price competition with itself. A better strategy would be introducing a damaged good (Switch Lite), or planned obsolescence (Switch 2).

Should the Nikkei and Bloomberg reporting of FY24 production ramp-up be accurate, for the reasons above I doubt that a price cut would be its main driving force. I'm not certain that a new model is part of that forecast though. Who knows, perhaps Nintendo is banking on the Mario movie to generate a large new cohort.

P.S.: The numbers above are for illustration purposes only. I also did the calculations quickly, so let me know if you spot any errors.

 

[mjayer]

Wasn’t the F-Zero rumour a remaster of F-Zero GX?

I believe that was (negative) speculation, the rumor was more just vague hints about a f-zero game in the works. And I believe we got the same kind of rumor from multiple sources.

 

[AlwaysOnDeck]

Some have theorized that Nintendo might reduce the prices of all models by $50 to stimulate the sales.

Expecting a $50 price cut is fan fiction

If/when a price cut happens, it'll be $20, each. $179 Switch Lite, $279 Switch, no price cut on OLED in 2023.