StarTopic Future Nintendo Hardware & Technology Speculation & Discussion |ST| (New Staff Post, Please read)

[Deleted member 2]

Actually the best thing that could happen to Pokémon

You know Nintendo doesn't make Pokémon, right? And what you're asking "why not" about would mean literally years of no first party Nintendo titles?

 

[ILikeFeet]

The only anime thing about this is the skybox and the character models.

it's pretty in line with anime-styled games still. they don't do cel-shaded environments, just the characters

 

[Multiverse]

The only anime thing about this is the skybox and the character models.

it's pretty in line with anime-styled games still. they don't do cel-shaded environments, just the characters

It looks like one of those paint overs/photo filters. Ironically, anime studios have shifted to that method for backgrounds. It’s just cheaper.

Pretty much the same as stock assets in gaming terms lol.

 

[RennanNT]

The amount needed depends on the type and the target resolution. Since it is a device targeting 720-1080p internally(?), I guess if you want a good saving or reduction of Bandwidth requirements an 8-16MB SLC can work…?

To be clear, that discussion was about using 24GB at 68GB/s instead of 12GB at 102+GB/s and using the extra RAM as a calculation results to save some bandwidth (and that the cached results would be smaller than the inputs required to recalculate).

I'm not well versed in that type of caching, so I'll refrain on commenting there. But since the cache on chip is small and you can't put all games assets there, there's still a lot of other things besides those results which needs to be transferred and they have to reach on time if you don't want stutters.

So I don't see those saving compensating 32+GB/s loss in bandwidth. Specially so when, as with any cache, the savings only happens sometimes.

 

[SoldierDelta]

it's pretty in line with anime-styled games still. they don't do cel-shaded environments, just the characters

Considering how Xenoblade and Fire Emblem also take this approach, yep.

 

[Alovon11]

I don't think this was mentioned (?), but another reason for Nintendo to launch outside the holiday season is the following:
Whether they launch during the holidays or not, they will initially sell out. But by launching outside the holiday season, far more of the purchases will be by the less casual crowd, consisting of the early adopters, and these buy more games per console. So by launching outside the holiday season, Nintendo will be in a superior revenue position right from the start, and any game coming in that period has an even better chance of succeeding (+this gives more games more room to breath).

All imo of course.

Not only that, but I feel VGC's Reasoning for 2H24 is....inherently flawed, especially from the production angle.

VGC states that the reason for 2H24 is Nintendo is

"likely to release new hardware during the second half of 2024, to ensure that it has ample stock available on day one and to avoid the kind of shortages seen with PlayStation 5 and Xbox Series X|S"

This just...doesn't make any sense at all to be perfectly honest. Primarily due to storage costs as under this interpretation, and honestly based on the leaks themselves indicating status of the devkits. Full Production should be starting in October/November, or at the very latest January/February.

The arguing for this is...just look at the Original Switch's Devkit to Production to Release Cycle.

SDEV (The name for the OG switch devkits) was seemingly handed out to the "Final Rung" of pre-launch Third parties in July 2016 (as that was when the devkit dam burst for OG Switch with that Eurogamer Article and the images of the Devkits...and also Sonic team announcing Project Sonic 2017 for NX. An event I feel any dev on that list of pre-launch devs to get devkits would be annoyed at occurring before they got devkits)

While some may argue that Switch 2 may be longer than Switch 1 devkit -> full production becuase the latter was "Rushed" (IMHO, that's a bit silly. Yeah it was quick....but still around the same time/if longer than the "Last round" of Third parties that got PS5 Devkits pre-launch. (Bloober Team saying they got their devkits in March 2020, so 8mos. The "Last round" of pre-launch SDEV Switch unit rollout likely starting somewhere in May-July 2016), and PS5 had the advantage of being B/C so patches/methods of enhancement without a final devkit could be done far easier and closer to launch for the PS5's case.

And in my honest opinion...I don't see why Nintendo would let Switch 2 stay unannounced for that long now that we seemingly are in that final phase of third party rollout for devkits. VGC and NecroLipe already got leaks which I assume are from devs in that lattermost part of the list of devs to get systems pre-launch. So Full Production likely is going to start in October considering the timeline is lining up there. Or at the absolute latest, January. And in either case, that puts it well before Holiday following common Full Production -> Launch Timeline patterns

Rolling this back to VGC....the reason why Nintendo "Releasing in 2H24 to ensure supply" is kidna dumb is...that kind of isn't how production works. You can't start producing nearly/over a year from launch and just hold onto those units with no qualms. Storage gets expensive fast. And not to mention it adds another major point of unpredictability in cost. Storage isn't like TSMC or Samsung where you can go to the singular entity and barter to some extent. Storage for consoles requires talking with multiple warehouse owners, securing them, making sure storage is safe for them, paying all that and continuously paying more and more as the warehouse owner(s) have to decline other deals for storage to keep holding onto your product.

So if Nintendo started production, heck, even in January for a Holiday launch, or even a September launch. They likely would be set up to pay what would be 6 months of sales just covering the storage costs. So yeah they'd hypothetically have double the units...but unless there is a absolutely insane markup on Switch 2 vs it's production cost, they'd need to sell every single one of those units to make the pre-launch production period cost back by any appreciable extent versus breaking even, or worse, a loss.

 

[Sol]

Only thing I'm certain about is that MS console library is 100% forward compatible. Less certain that Sony and Nintendo will feature more than one generation of BC.

Yeah, I'm seriously considering buying all future third party games on Steam instead of my PS5 (which is my main gaming system). I just can't trust that Sony will honor my digital purchases going forward on the PS6 and future hardware.

I feel the same about Nintendo. Switch 2 will undoubtedly have BC with our Switch games. But the system after the Switch 2? Doubtful.

Meanwhile I know I'll be able to build a new PC in 2040 and still be able to play any game in my current Steam library.

 

[Giancarlo]

Being on Unreal doesn't mean the art style has to look

What engine does nintendo use in most of their games anyway? Is it a in house engine?

is a engine know as Module system

 

[Giancarlo]

Not only that, but I feel VGC's Reasoning for 2H24 is....inherently flawed, especially from the production angle.

VGC states that the reason for 2H24 is Nintendo is


This just...doesn't make any sense at all to be perfectly honest. Primarily due to storage costs as under this interpretation, and honestly based on the leaks themselves indicating status of the devkits. Full Production should be starting in October/November, or at the very latest January/February.

The arguing for this is...just look at the Original Switch's Devkit to Production to Release Cycle.

SDEV (The name for the OG switch devkits) was seemingly handed out to the "Final Rung" of pre-launch Third parties in July 2016 (as that was when the devkit dam burst for OG Switch with that Eurogamer Article and the images of the Devkits...and also Sonic team announcing Project Sonic 2017 for NX. An event I feel any dev on that list of pre-launch devs to get devkits would be annoyed at occurring before they got devkits)

While some may argue that Switch 2 may be longer than Switch 1 devkit -> full production becuase the latter was "Rushed" (IMHO, that's a bit silly. Yeah it was quick....but still around the same time/if longer than the "Last round" of Third parties that got PS5 Devkits pre-launch. (Bloober Team saying they got their devkits in March 2020, so 8mos. The "Last round" of pre-launch SDEV Switch unit rollout likely starting somewhere in May-July 2016), and PS5 had the advantage of being B/C so patches/methods of enhancement without a final devkit could be done far easier and closer to launch for the PS5's case.

And in my honest opinion...I don't see why Nintendo would let Switch 2 stay unannounced for that long now that we seemingly are in that final phase of third party rollout for devkits. VGC and NecroLipe already got leaks which I assume are from devs in that lattermost part of the list of devs to get systems pre-launch. So Full Production likely is going to start in October considering the timeline is lining up there. Or at the absolute latest, January. And in either case, that puts it well before Holiday following common Full Production -> Launch Timeline patterns

Rolling this back to VGC....the reason why Nintendo "Releasing in 2H24 to ensure supply" is kidna dumb is...that kind of isn't how production works. You can't start producing nearly/over a year from launch and just hold onto those units with no qualms. Storage gets expensive fast. And not to mention it adds another major point of unpredictability in cost. Storage isn't like TSMC or Samsung where you can go to the singular entity and barter to some extent. Storage for consoles requires talking with multiple warehouse owners, securing them, making sure storage is safe for them, paying all that and continuously paying more and more as the warehouse owner(s) have to decline other deals for storage to keep holding onto your product.

So if Nintendo started production, heck, even in January for a Holiday launch, or even a September launch. They likely would be set up to pay what would be 6 months of sales just covering the storage costs. So yeah they'd hypothetically have double the units...but unless there is a absolutely insane markup on Switch 2 vs it's production cost, they'd need to sell every single one of those units to make the pre-launch production period cost back by any appreciable extent versus breaking even, or worse, a loss.

VGC report on the dev kit of Switch sucessor, dont make it clearly, if Nintendo has sent a early or a finalized dev kit for they next console, is just important third party partners to Nintendo, already have the Switch sucessor dev kit, does Capcom, Ubisoft, Square Enix, have a early or finalized dev kit for Switch sucessor? this will determine if Switch sucessor is a second half/holiday 2024 launch or a spring 2025 launch as VGC claim in it report.

 

[zdfg454]

The game looks great because the devs made a good job.
You will see great games and duds on every engine.

Epic is doing some tech edge stuff, such as Lumen and Nanite, but these are mostly targeted at high-end machines.
I do agree the invest a lot in R&D, but so does Nintendo.
Technically, there's nothing Nintendo cannot replicate internally, and they probably will in some form but optimized for their use-cases.

The strength of Unreal is mostly its ecosystem; SDK, editors, tooling, tutorials and a large marketplace.
This and its many abstractions make it easier to learn than private engines, which are usually closer to the metal and have little documentation.
That's probably why Nintendo uses it when partnering with external studios, it's easier to get it up and running and easier to hire and train employees.

I'm not sure what Epic's licensing model is right now, but I think it's probably around 20-30% of the revenue for games that sell well.
(I believe there's a top limit after which you don't pay Epic anymore. I believe The Witcher 3 hit it at 50M copies, but I'm not sure).
In Nintendo's case, with how many copies they're selling, they would probably have to pay Epic hundreds of millions annually, if not more, enough to fund multiple internal engines.

In Nintendo's case, not only do they save money developing their own engine, they can optimize it a lot better for what they need.
And the engine team is or will be in the same buildings as the EPD teams, in direct contact and giving direct support.
Epic, with the language barrier from an opposite timezone, could never support Nintendo on the same level.

The following video is great high-level introduction into how BotW in rendered.
They're doing very custom stuff that I believe would be more difficult to replicate in other engines.
I'm not even sure Unreal could run a game like BotW on three A57 CPUs, or have that level of customization in its shaders.
This clearly shows that Nintendo has some great rendering experts that go very low-level.
They would probably completely override Unreal's rendering stack if they were to use it, negating its very purpose.

Frequently Asked Questions - Unreal Engine

Thanks for considering Unreal Engine for your development needs. Here you will find a list of common questions answered to make informed decisions with little guesswork.

www.unrealengine.com

According to the official FAQ
Unreal Engine takes 5% of revenue when a game generates over $1 million in sales.
This means that up to $1 million, Unreal Engine is free to use, and the 5% royalty can be reduced or eliminated under customized license terms with Epic Games.
With royalties this low it's no wonder that even the big game companies are abandoning their own game engines and moving to Unreal.
It's expensive and time-consuming for a game company to build their own game engine, and it's devastating if the features don't meet expectations.
Using Unreal for game development is a cheap and safe choice.

 

[Joy-Boy]

Fallout 4 was scheduled to have a Next Gen patch this year, but it looks like it isnt going to make it for a bit.

Sure would be nice if there was a whole new system launching next year that they could release it on

I doubt Microsoft will do it, but Skyrim made buttloads launching with Switch and Fallout 4 already isnt exclusive so maybe they'll allow it

 

[Concernt]

Fallout 4 was scheduled to have a Next Gen patch this year, but it looks like it isnt going to make it for a bit.

Sure would be nice if there was a whole new system launching next year that they could release it on

I doubt Microsoft will do it, but Skyrim made buttloads launching with Switch and Fallout 4 already isnt exclusive so maybe they'll allow it

Pete Hines really needs to work on his tone...

 

[MusicProposition]

Yes which applies to every other product on planet earth. Consumers are clearly not that fickle. Gaming consoles rely on their software schedule, these are gaming consoles so games is their priority. When platforms lack unappealing games despite great features we see them suffer. As long as Nintendo provides their system selling ip, consumers will consume.

It certainly is, games sell systems. As long as Nintendo provides system selling games such as Mario, Zelda, and Pokemon, consumers will consume. The biggest difference between these consoles has always been the library behind them with every other feature being secondary. None of Nintendo's past platforms are comparable to this because they all had to compete for Nintendo's resources (games). It takes money and time to produce games, which is why those home consoles suffered from droughts and a lack of system selling games. Nintendo's handhelds have always had a consistent high selling franchise exclusive to them (Pokemon), that's why Nintendos handhelds are so consistent. Switch is about to instantly become Nintendo's highest selling platform due to it being the biggest beneficiary of having Nintendo's undivided attention.

Both the N64 and the GameCube had Mario, Zelda an Pokémon (not mainline) games and I really wouldn’t consider those consoles a success. I do think it’s not as easy as just providing games, since the market and the consumer‘s needs can always change.

 

[MusicProposition]

Fallout 4 was scheduled to have a Next Gen patch this year, but it looks like it isnt going to make it for a bit.

Sure would be nice if there was a whole new system launching next year that they could release it on

I doubt Microsoft will do it, but Skyrim made buttloads launching with Switch and Fallout 4 already isnt exclusive so maybe they'll allow it

keep expectations for 3rd Party ports low…

 

[Concernt]

keep expectations for 3rd Party ports low…

The worst possible thing to have: hope.

 

[MuchLonk]

Switch 3rd party outputs were not bad if you compare with previous console imo.
I can see the same fate for the Switch 2

 

[Dekuman]

Switch 3rd party outputs were not bad if you compare with previous console imo.
I can see the same fate for the Switch 2

should be better

The key is Nintendo to maintain the momentum (re. hardware sales and attachment to their ecosystem) which is why 2024 launch makes sense, even though understandably, some people feel its too soon givne how well OG Switch is still selling.

 

[ReyVGM]

keep expectations for 3rd Party ports low…

As a Nintendo fan in a post-SNES era? expectations are ALWAYS low.

 

[Deleted member 887]

To be clear, that discussion was about using 24GB at 68GB/s instead of 12GB at 102+GB/s and using the extra RAM as a calculation results to save some bandwidth (and that the cached results would be smaller than the inputs required to recalculate).

I'm not well versed in that type of caching, so I'll refrain on commenting there.

Cache misses operate by a power law. Every doubling of cache, in general, halves the number of misses.

In this case, we're talking about bandwidth dropping linearly. Speaking purely in terms of the cache, you're talking about doubling a cache's size, and at best 25% percent increase in cache effectiveness, for a 50% drop in cache speed. That means the things you are already caching would be half as fast.

Putting aside how important cache operations are in the context of frame rendering, even purely as cache, this is a huge loss. That's why every on-die caching system has L0, L1, L2, L3 - so they can add cache without dropping perf across the board.

In the real world, the situation would be even worse. In memory caching operations usually use hash tables. Hash tables have access times that reduce as memory size increases. But in the real world, there is a limit to how much caching would help. Devs would likely not cache compute operations, but instead cache storage loads.

Loads don't happen every frame, so while you could eliminate traversal and loading stutter, you would do so by greatly reducing visuals in every frame. One way to think of memory bandwidth is "how much data can I use each frame?" At 102GB/s, that's something like 1.6 GB of data per frame, plenty of high res textures, lots of geometry. Halve that to 800MB, and you need significant reductions in texture quality and geometry to account for it.

 

[Pokemaniac]

What engine does nintendo use in most of their games anyway? Is it a in house engine?

At this point it's mostly per studio. EPD has mostly standardized on ModuleSystem, and Monolith, Retro, and NLG all have their own tech. Outside studios use a mix of their own engines, standard ones like Unity or Unreal, and sometimes NintendoWare Bezel, which appears to be a generic engine that third parties can license from Nintendo.

 

[Cybergatuno]

Frequently Asked Questions - Unreal Engine

Thanks for considering Unreal Engine for your development needs. Here you will find a list of common questions answered to make informed decisions with little guesswork.

www.unrealengine.com

According to the official FAQ
Unreal Engine takes 5% of revenue when a game generates over $1 million in sales.
This means that up to $1 million, Unreal Engine is free to use, and the 5% royalty can be reduced or eliminated under customized license terms with Epic Games.
With royalties this low it's no wonder that even the big game companies are abandoning their own game engines and moving to Unreal.
It's expensive and time-consuming for a game company to build their own game engine, and it's devastating if the features don't meet expectations.
Using Unreal for game development is a cheap and safe choice.

OK, it's not as bad as I thought.

In Nintendo's case, that still means $250-500K every year. It's not "cheap" and it doesn't even include support services.

For small to mid-sized studios, it absolutely makes sense. They can't re-invent the wheel for every little aspect of a game.
But for some reason most big publishers/developers to have their own engines and some are even looking into licensing them, because they are also very good.
The bigger a company is, the more investing in their existing engine "is a cheap and safe choice".

To get the most out of Unreal, you need rendering/GPU experts that understand how everything translates to the lowest levels and are able to inspect and optimize it.
Big companies have those anyway, so most of the investment for a private engine is already being made.

Game engines are not that important, development talent is.
You load your shaders and assets on top on existing standards, OpenGL or Vulkan or whatever, the engine abstracts it, and you setup your command list for the rendering loop.
Then you have runtime libraries to handle the game state, sound, animations, etc. and development-time editors and tooling.
There are open-source implementations of almost anything you would need and big companies probably already have multiple implementations of those.

Unreal just puts everything in a nice modular package with good documentation and good default out-of the-box implementations.

 

[GrandDemand]

should be better

The key is Nintendo to maintain the momentum (re. hardware sales and attachment to their ecosystem) which is why 2024 launch makes sense, even though understandably, some people feel its too soon givne how well OG Switch is still selling.

I think sales of NX1 will fall off quicker than some people believe, they got a huge bump from ToTK IIRC. Will be interesting to see Q3 results as that would represent typical sales volume (unlike Q4 Holiday shopping season)

 

[Concernt]

Switch 3rd party outputs were not bad if you compare with previous console imo.
I can see the same fate for the Switch 2

Switch 2 will be cheaper and easier to port to from Gen 9 than 1 was from Gen 8, AND it'll be EVEN EASIER to bring Gen 8 (and thus cross gen games) to Switch 2. This whole thing stinks of pessimism. It'll be fine. As well supported as Xbox? Sure, probably not? "Prepare for disappointment" levels of non-support? Buddy, this isn't Wii U.

 

[SoldierDelta]

At this point it's mostly per studio. EPD has mostly standardized on ModuleSystem, and Monolith, Retro, and NLG all have their own tech. Outside studios use a mix of their own engines, standard ones like Unity or Unreal, and sometimes NintendoWare Bezel, which appears to be a generic engine that third parties can license from Nintendo.

It's worth pointing out, however obvious it may seem, that several studios that do support work (best example being Monolith on TotK) have experience on various other Nintendo game engines. While there isn't a unifying engine tying them together (as far as I know, I'm just a dude with a Sena PFP not an insider), it's not like studio's experience is mutually exclusive from each other.

As a guess for the next-generation (would like some help clarifying this btw), Nintendo will probably get the less experienced devs to work on less-demanding projects (whether support work or remake/remasters) just to get experience on 8th gen hardware before making meatier games. This would allow devs time to update/fix their engines to work on the more demanding hardware. This is the first time Nintendo has had a hardware upgrade in 10 years, so it behoves them to actually test waters appropriately.

 

[ILikeFeet]

OK, it's not as bad as I thought.

In Nintendo's case, that still means $250-500K every year. It's not "cheap" and it doesn't even include support services.

For small to mid-sized studios, it absolutely makes sense. They can't re-invent the wheel for every little aspect of a game.
But for some reason most big publishers/developers to have their own engines and some are even looking into licensing them, because they are also very good.
The bigger a company is, the more investing in their existing engine "is a cheap and safe choice".

To get the most out of Unreal, you need rendering/GPU experts that understand how everything translates to the lowest levels and are able to inspect and optimize it.
Big companies have those anyway, so most of the investment for a private engine is already being made.

Game engines are not that important, development talent is.
You load your shaders and assets on top on existing standards, OpenGL or Vulkan or whatever, the engine abstracts it, and you setup your command list for the rendering loop.
Then you have runtime libraries to handle the game state, sound, animations, etc. and development-time editors and tooling.
There are open-source implementations of almost anything you would need and big companies probably already have multiple implementations of those.

Unreal just puts everything in a nice modular package with good documentation and good default out-of the-box implementations.

as always, Nitnendo wouldn't pay that to Epic. large companies pay up front (or something completely different) that covers many games and custom support. for the talk of major studios moving away from in-house engines, this is a big sticking point for them

• There are also options for custom licenses that can include premium support; private training; negotiated terms for lower royalties, no royalties, or a different basis for royalty negotiation; and more. Contact us to inquire about a custom license for either games or non-games use.

 

[MP!]

Switch 2 will be cheaper and easier to port to from Gen 9 than 1 was from Gen 8, AND it'll be EVEN EASIER to bring Gen 8 (and thus cross gen games) to Switch 2. This whole thing stinks of pessimism. It'll be fine. As well supported as Xbox? Sure, probably not? "Prepare for disappointment" levels of non-support? Buddy, this isn't Wii U.

Pretty sure that’s what muchlonk is saying

 

[SoldierDelta]

as always, Nitnendo wouldn't pay that to Epic. large companies pay up front (or something completely different) that covers many games and custom support. for the talk of major studios moving away from in-house engines, this is a big sticking point for them

This is very interesting to learn about. Knowing Nintendo (Mr. "We decide our fate" themselves) they probably would try everything they can to avoid insane royalties to Epic via whatever means necessary.
For smaller games this doesn't make as much sense, and would explain why they decided to use it for, say, Pikmin 4. However, for games like Tears of the Kingdom, it's worth the extra development costs to make/modify their own engine instead of just licensing UE.

If I made any mistake please correct me, I'm not too knowledgeable about this.

 

[Angel Whispers]

keep expectations for 3rd Party ports low…

Please don’t do this.

 

[SoldierDelta]

Switch 2 will be cheaper and easier to port to from Gen 9 than 1 was from Gen 8, AND it'll be EVEN EASIER to bring Gen 8 (and thus cross gen games) to Switch 2. This whole thing stinks of pessimism. It'll be fine. As well supported as Xbox? Sure, probably not? "Prepare for disappointment" levels of non-support? Buddy, this isn't Wii U.

Lonk pretty much said this.

The Switch's early life did kinda show that devs weren't that interested in the Switch (see the Wii U as to why), but once it sold more than the population of Mexico that changed their minds.
I honestly think that the third-party support for the Switch 2 day one is going to be fucking nuts. Hell, I wouldn't be all too surprised if Take-Two/Rockstar port GTA6 within the first year.

 

[lexony]

For graphics, UE is just incredible and Pikmin 4 looks phenomenal.

If they can do everything with UE, why not?

The look of Games aren‘t (only) defined because of their engine. You can make a highly realistic good looking game also in Unity if you really wanted to. If Nintendo wants to use Unreal for some of their games, fine it‘s a great tool especially when you‘re looking after realistic graphics, lighting and all that shit. Doing a game like Ring Fit? Making the next Mario title? Nah Unreal makes things just unnecessarily more complicated.

Edit: At the End of the day a dev team should use the engine where they feel most comfortable in to develop the game they want to achieve for a certain platform. Nintendo seem very open to that considering their extensive use of different internal and external engines for their games and this is only a good thing that they should keep. Otherwise you end up with something like EA had with struggling for years with Frostbite or obviously Gamefreak who are seemingly not able to further develop their engine for a modern platform.

 

[ILikeFeet]

This is very interesting to learn about. Knowing Nintendo (Mr. "We decide our fate" themselves) they probably would try everything they can to avoid insane royalties to Epic via whatever means necessary.
For smaller games this doesn't make as much sense, and would explain why they decided to use it for, say, Pikmin 4. However, for games like Tears of the Kingdom, it's worth the extra development costs to make/modify their own engine instead of just licensing UE.

If I made any mistake please correct me, I'm not too knowledgeable about this.

these custom licenses cover many games. for Square Enix, for instance, they pay up front so they don't have to pay royalties for any of their games during the period.

this is an older post and has probably expired, but it worked out very well for them so it wouldn't be surprising for them to make another agreement

Multi-title deal provides Square Enix developers with full access to Unreal Engine 3 and Unreal Engine 4 technology across its studios​

CARY, N.C. – Epic Games, Inc. announced today an agreement with Square Enix Co. Ltd. granting Square Enix a group-wide license to develop multiple games with Epic’s award-winning Unreal Engine 3 and its breakthrough Unreal Engine 4 technology.

Epic Games Japan, which is providing local support, worked directly with Square Enix on the new license. This sweeping agreement will supplement Square Enix’s proprietary software and ensure the studio is armed with the most powerful game development tools and technologies available across all major platforms.

“Epic has a strong relationship with Square Enix, and it’s an honor to provide their creative and technical talent with the best game technology available for licensing,” said Taka Kawasaki, territory manager of Epic Games Japan. “We look forward to supporting their developers with world-class tools for years to come.”

“Demons’ Score” for iOS and Android, developed by iNiS and published by Square Enix, was created with Unreal
Engine 3.

Square Enix Signs Long-Term Unreal Engine Licensing Agreement

CARY, N.C. – Epic Games, Inc. announced today an agreement with Square Enix Co. Ltd. granting Square Enix a group-wide license to develop multiple games with Epic’s award-winning Unreal Engine 3 and its breakthrough Unreal Engine 4 technology.

www.unrealengine.com

 

[MusicProposition]

Please don’t do this.

what do you mean?

 

[Pokemaniac]

It's worth pointing out, however obvious it may seem, that several studios that do support work (best example being Monolith on TotK) have experience on various other Nintendo game engines. While there isn't a unifying engine tying them together (as far as I know, I'm just a dude with a Sena PFP not an insider), it's not like studio's experience is mutually exclusive from each other.

As a guess for the next-generation (would like some help clarifying this btw), Nintendo will probably get the less experienced devs to work on less-demanding projects (whether support work or remake/remasters) just to get experience on 8th gen hardware before making meatier games. This would allow devs time to update/fix their engines to work on the more demanding hardware. This is the first time Nintendo has had a hardware upgrade in 10 years, so it behoves them to actually test waters appropriately.

Realistically, I think the less demanding portions of Nintendo's lineup are going to take a while to move on from Switch 1. The Switch's continued success doesn't mean Nintendo are going to delay its successor, but it does mean that Nintendo isn't going to drop it the second they have something else to replace it. These titles will likely be the bulk of Nintendo's "cross-gen" titles.

 

[LiC]

Hey guys, looks like the Pokemon Rumor was right about the new testral mechanic

I'd like the remind everybody that all the Pokémon leak post purported to say about this mechanic was "Special Terastal appearance" and "Some Pokémon will acquire new Terrastal forms." Not much of a prediction.

Also, even if they did have info beyond that week's Presents, that still doesn't prove they know anything about new hardware. Their story about being a programmer and their (quite possibly faked) translated Japanese post are still incredibly suspicious.

 

[Ojoloco]

I'd like the remind everybody that all the Pokémon leak post purported to say about this mechanic was "Special Terastal appearance" and "Some Pokémon will acquire new Terrastal forms." Not much of a prediction.

Also, even if they did have info beyond that week's Presents, that still doesn't prove they know anything about new hardware. Their story about being a programmer and their (quite possibly faked) translated Japanese post are still incredibly suspicious.

As I said in a previous post, the supposed leak mentioned things that had not been mentioned in any previous leak or rumor, and that detail was the only thing that was missing to be confirmed beyond the 2024 date of the dlc2 and the next gen patch .

I think it's okay to take all the rumors with a grain of salt, but the supposed pokemon leak is equally curious and interesting.

 

[Z0m3le]

The SoC needs to render a frame every 33ms or lower. Every data it requires to render that frame (objects, their position and surfaces, textures, light sources, etc) needs to reach it in time. Cache is really expensive, so there's only a few MBs available and thus you need a good transfer speed from the RAM to the APU or the later won't be able to do it's job in time.

I'm not much familiar with how much bandwidth you can save with caching results in RAM, but there's only so much you can do there. And 100~135 GB/s isn't that much to begin with.


Maybe I wasn't clear in my previous post, but the main benefit isn't the multichannel in itself. If the APU has 128-bit width, you're ideally using a total of 128-bit memory module(s). A 128-bit module has the same bandwidth as 2x 64-bit module. The main problem with using the one you linked alone is that it is 64-bit, not that it wouldn't have dual channel.

That's what the PS5 is doing. It's filling the 256-bit bus with 8 modules of 32-bit 2GB GDDR6 for 448 GB/s total. This is more about what is available in the market though.


It's not about the number of channels, just the sum of all modules bandwidth (and the cap of the chip).


Drake has a 128-bit bus width. There's no point in using RAM beyond 128-bit, so a 3 modules config means 64-bit + 32-bit + 32-bit.

If Drake was 192-bit, like the Series S IIRC, then that would work. But a bigger bus also has impact on power consumption (and they using 128-bit is already a signal they don't think 64-bit had enough bandwidth) and more modules require more space in the motherboard (which is a premium for a tablet).

All in all, it's a balancing act. You don't want to starve the APU or lack in total RAM, but Nintendo/NVidia also have to worry about not going overkill and spend too much money/space/battery life on RAM instead of some other aspect in more need.

IIRC the RTX 3050 has 160GB/s? the RTX 2050 has 112GB/s, and these are similar class GPUs to watch Drake will have, especially if Drake relies on mixed precision like TX1 did in the Switch, that would give a fairly comparable performance to those cards, especially with the larger RAM available in Drake over those GPUs.

I do think Drake will out perform our expectations, this chip is twice as large as we thought it would be, and it's not going to be throttled to the ground simply because a chip half it's size would perform the same, making the chip nonsense, and Nvidia is the designer of the chip, Nintendo just agrees to performance targets, Nvidia is the one who gets them there.

 

[Deleted member 887]

As a guess for the next-generation (would like some help clarifying this btw), Nintendo will probably get the less experienced devs to work on less-demanding projects

internally, you’ll put inexperienced devs under experienced devs, and you’ll concentrate engine work in a tools/support team which allows maximum knowledge sharing.

When it comes to external devs, I think the driving factor will be Nintendo’s relationship with them. Which does tend to correlate with the team’s age and global experience, but isn’t 1:1.

 

[Cybergatuno]

...

 

[thesillyfellaelde]

EA had with struggling for years with Frostbite or obviously Gamefreak who are seemingly not able to further develop their engine for a modern platform.

Yeah EA forcing their studios to only use the Frostbite engine for ever game is a good example of how choosing an engine designed for specific types of games makes getting other types of games working on it sometimes very painful. I see a lot of people wanting Pokemon on Unreal Engine and I'd be all for that, but it's no guarentee that it'd run smoothly on Unreal with how the Pokemon games are becoming more "open world". There's not many open world games on Unreal Engine 4 to my knowledge, but the ones that do have big open landscapes are kinda notorious for being hell to run already (Ark Survival Evolved, PUBG)

 

[RennanNT]

IIRC the RTX 3050 has 160GB/s? the RTX 2050 has even less,

Yeah, the weaker the card, the lower is the resolution/assets it's targeting and thus it's bandwidth needs are lower as well.

Someone made a comparison and Nvidia is quite consistent with giving 25~30 GB/s per TF on most of their Ampere cards. Once you account some bandwidth for the CPU, that would make 102GB/s about right for a 3~3.2TF Drake. Not sure if use mixed precision would require higher bandwidth though.

Meanwhile 68GB/s would likely means Drake will be bandwidth starved or target something which 8SM would accomplish for cheaper.

 

[SoldierDelta]

internally, you’ll put inexperienced devs under experienced devs, and you’ll concentrate engine work in a tools/support team which allows maximum knowledge sharing.

When it comes to external devs, I think the driving factor will be Nintendo’s relationship with them. Which does tend to correlate with the team’s age and global experience, but isn’t 1:1.

I was moreso thinking about dev's experience with higher-end hardware (which is probably my fault, my phrasing is bloody awful). With the exception of Nintendo EPD and Monolith Soft whom already make very high-end games, I feel like a lot of devs are going to be heavily effected by the new hardware jump and expectations.

The massive hiring is also a problem, Nintendo is expanding their internal teams by a comical amount and are planning on opening more in the future (praying for more Monolith Soft teams fr fr), but i'm more concerned by the present problem of bigger hardware and workloads.

 

[LiC]

As I said in a previous post, the supposed leak mentioned things that had not been mentioned in any previous leak or rumor, and that detail was the only thing that was missing to be confirmed beyond the 2024 date of the dlc2 and the next gen patch .

I think it's okay to take all the rumors with a grain of salt, but the supposed pokemon leak is equally curious and interesting.

There were two unique things in their post: "Special Terastal appearance" and the claims about new hardware. Everything else, 99% of the posts, was information from the Pokémon Presents that aired within a week. Pokémon Presents contents are often leaked early on /vp/, so that's not that special. The question is what proves they had information beyond the Presents, to potentially make their claims about hardware credible.

 

[SoldierDelta]

Realistically, I think the less demanding portions of Nintendo's lineup are going to take a while to move on from Switch 1. The Switch's continued success doesn't mean Nintendo are going to delay its successor, but it does mean that Nintendo isn't going to drop it the second they have something else to replace it. These titles will likely be the bulk of Nintendo's "cross-gen" titles.

Ah, I kinda forgot about Cross-gen for a second.
I feel like we'll get a large bulk of them, with very few studios (most notably EPD, Monolith Soft and maybe Retro) actually focusing on next-gen exclusives. If this is what it takes to get teams to get fully used to the bigger workloads, I'm for it.

 

[Ojoloco]

There were two unique things in their post: "Special Terastal appearance" and the claims about new hardware. Everything else, 99% of the posts, was information from the Pokémon Presents that aired within a week. Pokémon Presents contents are often leaked early on /vp/, so that's not that special. The question is what proves they had information beyond the Presents, to potentially make their claims about hardware credible.

The point is that it is a curious leak that is hitting many things.
In fact, the fact that part of the things were presented in a Presents does not change the matter much, since this leak happens much earlier and contains certain things that are not presents in that event.

Nor do I think it is necessary to emphasize multiple times that these leaks be taken with a graint of salt.

 

[thesillyfellaelde]

Ah, I kinda forgot about Cross-gen for a second.
I feel like we'll get a large bulk of them, with very few studios (most notably EPD, Monolith Soft and maybe Retro) actually focusing on next-gen exclusives. If this is what it takes to get teams to get fully used to the bigger workloads, I'm for it.

Surpriiiise, it's actually the last few Wii U games still stuck on the Wii U being ported
One thing I wonder is how long they're gonna make cross-gen games for the Switch and Switch 2, would making the next Mario Kart be available on both Switch and Switch 2 be good?

 

[Giancarlo]

I think sales of NX1 will fall off quicker than some people believe, they got a huge bump from ToTK IIRC. Will be interesting to see Q3 results as that would represent typical sales volume (unlike Q4 Holiday shopping season)

i higly doubt this, Super Mario Bros Wonder together with Tears of the Kingdom, will make the Switch sales explode, just look at Tears of the Kingdom that managed to sell nearly 19 milions units in just 50 days, Super Mario Bros Wonder will be the best selling game of the franchise.

 

[SoldierDelta]

Surpriiiise, it's actually the last few Wii U games still stuck on the Wii U being ported
One thing I wonder is how long they're gonna make cross-gen games for the Switch and Switch 2, would making the next Mario Kart be available on both Switch and Switch 2 be good?

For what it's worth, Xenoblade X is probably going to be Switch 2 exclusive due to the fact it only just worked on Wii U.

As for how long... uhhh... probably 2 years of support? Maybe 3 depending on demand at that point?

 

[Giancarlo]

Surpriiiise, it's actually the last few Wii U games still stuck on the Wii U being ported
One thing I wonder is how long they're gonna make cross-gen games for the Switch and Switch 2, would making the next Mario Kart be available on both Switch and Switch 2 be good?

expected games to be Switch sucessor games: next 3D Mario, Monolith Soft action RPG/Xenoblade Chronicles 4

expected games to be cross-gen: Metroid Prime 4, the rumored 2D/3D Donkey Kong game

 

[GrandDemand]

i higly doubt this, Super Mario Bros Wonder together with Tears of the Kingdom, will make the Switch sales explode, just look at Tears of the Kingdom that managed to sell nearly 19 milions units in just 50 days, Super Mario Bros Wonder will be the best selling game of the franchise.

Forgot about Wonder! And that totally wouldn't surprise me 2D Mario always sells excellently

 

[Concernt]

.