• Hey everyone, staff have documented a list of banned content and subject matter that we feel are not consistent with site values, and don't make sense to host discussion of on Famiboards. This list (and the relevant reasoning per item) is viewable here.
  • Do you have audio editing experience and want to help out with the Famiboards Discussion Club Podcast? If so, we're looking for help and would love to have you on the team! Just let us know in the Podcast Thread if you are interested!

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

If it’s August with no new real movement going into fall you can comfortably say it’s not gonna be this FY. That still gives us June or July for more info, but it’s coming down to the wire.
Coming down to the wire is a great way to put it, yeah. Pretty much sums it up this summer.
 
0
quad-core processors when the Orin chip with 1024 CUDA Cores has an octa-core one,
It also has a 6 core variant. The Quad vs Octo discussion isn't pessimism, it's asking technically whether octo core is a performance win over quad core, and if it isn't, wouldn't Nintendo take the higher yields/lower cost 4 core config?

If the goal is to release a revision that improves the existing library and opens up games to “enhanced” editions, then 4 cores which can run at higher clocks is the winner. Patching a game to take advantage of the faster cores is 4 lines of code. Re-engineering a game to take advantage of more cores is more complex. Even for exclusives arguably 4 faster cores is better than 8 slower cores.

If Nintendo is, instead, planning a successor that does backwards compat, then the cost/benefit analysis is very different. This isn’t copium.

Nintendo ran with the GameCube architecture for 3 generations. If Drake is a soon-to-release revision, then we should both expect that it will be cut down where it makes sense and that it represents large room for growth on a successor device where the existing library isn’t the primary concern.
 
0
The storage question isn't "because Nintendo" or copium either. It's about whether faster storage makes sense given the memory bandwidth constraints.

The big 4k twins use GDDR of some description and have a pretty decent Zen 2 CPU that can quickly decompress assets and throw them into memory. Hence the benefits of having fast NVME drives.

With the switch, its using LPDDR, which is much slower, again something we know, and will absolutely use a CPU with less grunt than the competition.

The only thing we have to compare to right now with a similar setup is the steam deck, and in the steam decks case the storage medium doesn't seem to impact load times. So is it worth Nintendo moving away from EMMC if the extra speed from going UFS is not going to manifest in practical terms? It also begs the question whether switch 2 will be capable of streaming assets like the PS5 even if it has UFS 4. Doesn't matter how fast the storage is if the memory bandwidth causes a traffic Jam.
 
For what it's worth, I'm suspecting that the Steam Deck is throttling the NVMe drive with very low power active states. Not that I can back that suspicion up... nor do people seriously review that sort of thing?
Of course, I also have the suspicion that the choice of NVMe SSD as storage for the higher tiers of that device is more for marketing than for practical utility...

Regarding the mystery of CPU grunt -> sequential reading speed, that is tough to find out, isn't it? We know empirically that the A57@1 ghz is the bottleneck with the Switch and that profiles raising the CPU clock exist for helping with loading. We don't know how much things have improved since the A57.
Anybody know how to read AndroBench scores? Are they basically just the speed with MB/s chopped off? Cause I'm looking at this (gotta scroll down a fair bit to get to AndroBench). The Snapdragon 865 has an A77 that bursts up to 2.84 ghz, so I'm assuming that's what gets the... 1,528 MB/s sequential read, I'm assuming?
If memory bandwidth is a concern for this benchmark, 6 gigs for LPDDR5 implies 64-bit.
 
At this rate I feel like 2022 is off the table with more and more certainty as the days pass. I know people got burned last year and people took a lot of heat so they may be more hesitant to speak this year, but I still feel like we’d be hearing something from others if this was a 2022 device. And I don’t think Mochizuki is deterred by blowback, he’s a business journalist, he doesn’t care what Twitter/forums think.

I’m not sure how far to apply this logic to Q1 2023…I’ve been pretty firmly on the “this FY” train but I have way less clue when we’d be hearing things if this was a March 2023 release. Maybe not yet. I still think in the next couple months if we’re in the same spot that doesn’t bode well. We’ll be 7-8 months out at that point.

They could easily announce the Drake
mid gen revision model in Sept and release it Oct/Nov
 
They could easily announce the Drake
mid gen revision model in Sept and release it Oct/Nov
Idk, I think a 4-8 week window for a device like this is cutting it way too close. Maybe 8 weeks I guess, I can’t remember what the Lite was, but no way it’s only a month in advance. Leaks for days.
 
They could easily announce the Drake
mid gen revision model in Sept and release it Oct/Nov

I think the issue here is that we'd expect a leak (or leaks) to start popping up very soon if things were coming this year.

I'm not sure if something happened after the dev kit leaks came about, but it's been quiet. Call me a fool, but I’m personally going to remain hopeful until late June
 
I think the issue here is that we'd expect a leak (or leaks) to start popping up very soon if things were coming this year.

I'm not sure if something happened after the dev kit leaks came about, but it's been quiet. Call me a fool, but I’m personally going to remain hopeful until late June
OLED was announced in July.

We never got the PS5 design unti reveal etc.

Things are way better being kept in secret nowadays…
 
For what it's worth, I'm suspecting that the Steam Deck is throttling the NVMe drive with very low power active states. Not that I can back that suspicion up... nor do people seriously review that sort of thing?
Of course, I also have the suspicion that the choice of NVMe SSD as storage for the higher tiers of that device is more for marketing than for practical utility...
Someone on Reddit did sequential speed benchmarks for the pre-installed 512 GB NVMe SSD, as well as the Sandisk Extreme Pro 1 TB microSD card, when inserted on the Steam Deck. The sequential read and write speeds for the pre-installed 512 GB NVMe SSD on the Steam Deck was 916 MB/s and 1.3 GB/s respectively. And the sequential read and write speeds for the Sandisk Extreme Pro 1 TB microSD card when inserted on the Steam Deck was 70.8 MB/s and 63.5 MB/s respectively.
 
Hmm, higher than I thought, but still lower power active state compared to what one would probably see on desktop. Should be sub 2 watts for the read. For the write, huh, for that speed, is that all SLC cache?
 
0
Someone on Reddit did sequential speed benchmarks for the pre-installed 512 GB NVMe SSD, as well as the Sandisk Extreme Pro 1 TB microSD card, when inserted on the Steam Deck. The sequential read and write speeds for the pre-installed 512 GB NVMe SSD on the Steam Deck was 916 MB/s and 1.3 GB/s respectively. And the sequential read and write speeds for the Sandisk Extreme Pro 1 TB microSD card when inserted on the Steam Deck was 70.8 MB/s and 63.5 MB/s respectively.
This is really interesting, I wonder why the actual load times of games don't differ much? Could it be game load times rely on random read rather than sequential?
 
This is really interesting, I wonder why the actual load times of games don't differ much? Could it be game load times rely on random read rather than sequential?
Asset decompression. On Switch, internal storage is 3x faster than SD card and over 10x the slowest cart, but the load time doesn't differ nearly as much. Then Nintendo create a profile to speed up loading times by increasing CPU clocks and even carts got benefits IIRC.

Basically, game assets are compressed to save space. Whenever they're needed, they go through the CPU, which uncompress them so the GPU can use them.

On PC this is even worse because of split RAM. The CPU decompress the file and put in the general use RAM and then have to send this to the RAM inside the graphic card.

This can be solved through dedicated hardware and OS help though. NVidia already implemented their solution, RTXIO, which allows the assets being decompressed directly by the GPU and use Microsoft's Direct Storage to avoid overheads. So, it should be feasible for them to achieve way faster on the successor.

I have no idea how much it would increase SoC price, storage price, power consumption, etc. Maybe it's within Nintendo's target budgets, maybe not.
 
I have no idea how much it would increase SoC price, storage price, power consumption, etc. Maybe it's within Nintendo's target budgets, maybe not.
the SoC wouldn't be affected as gpu decompression is a software function. Nvidia has been using it on linux for a while now
 
the SoC wouldn't be affected as gpu decompression is a software function. Nvidia has been using it on linux for a while now
Nice, but is this done on a GPU anywhere near the expected range for Drake? (honest question, I don't know).

I ask because I don't know how much software decompression would impact GPU performance compared to dedicated hardware (assuming streaming assets is a priority instead of just shrinking loading screens/tunnels, of course).
 
Nice, but is this done on a GPU anywhere near the expected range for Drake? (honest question, I don't know).

I ask because I don't know how much software decompression would impact GPU performance compared to dedicated hardware (assuming streaming assets is a priority instead of just shrinking loading screens/tunnels, of course).
we have no idea. DirectStorage isn't out yet (and thus RTXIO) so there's nothing to test it against. Forspoken will use DirectStorage but no GPU decompression, so that won't be much of a test
 
0
Asset decompression. On Switch, internal storage is 3x faster than SD card and over 10x the slowest cart, but the load time doesn't differ nearly as much. Then Nintendo create a profile to speed up loading times by increasing CPU clocks and even carts got benefits IIRC.

Basically, game assets are compressed to save space. Whenever they're needed, they go through the CPU, which uncompress them so the GPU can use them.

On PC this is even worse because of split RAM. The CPU decompress the file and put in the general use RAM and then have to send this to the RAM inside the graphic card.

This can be solved through dedicated hardware and OS help though. NVidia already implemented their solution, RTXIO, which allows the assets being decompressed directly by the GPU and use Microsoft's Direct Storage to avoid overheads. So, it should be feasible for them to achieve way faster on the successor.

I have no idea how much it would increase SoC price, storage price, power consumption, etc. Maybe it's within Nintendo's target budgets, maybe not.
Thanks for this, it was really informative. Nintendo could use RTXIO for shrinking loading times I imagine as the GPU shouldn't be doing too much at that time. This is really positive and this enhanced asset decompression speed could make faster storage such as UFS 3 more likely.

There is a reason I get hung up on this and its specifically because I am a huge resident evil fan. Got super hyped when RE1 and RE0 came out on switch as I could finally do speed runs on a portable device.

I'm not super fast, but RE1 best ending I can do in 80-90 minutes on the pc without mods. Switch version came out and whilst I was a little rusty the same run took me 140 minutes.

Almost twice the difference. This was 90% due to door loading times. My PC had a standard HDD and not a SSD at the time too whilst switch has solid state hardware. So to me this was just embarrassing.

Really hope Drake can fix this issue.
 
Thanks for this, it was really informative. Nintendo could use RTXIO for shrinking loading times I imagine as the GPU shouldn't be doing too much at that time. This is really positive and this enhanced asset decompression speed could make faster storage such as UFS 3 more likely.

There is a reason I get hung up on this and its specifically because I am a huge resident evil fan. Got super hyped when RE1 and RE0 came out on switch as I could finally do speed runs on a portable device.

I'm not super fast, but RE1 best ending I can do in 80-90 minutes on the pc without mods. Switch version came out and whilst I was a little rusty the same run took me 140 minutes.

Almost twice the difference. This was 90% due to door loading times. My PC had a standard HDD and not a SSD at the time too whilst switch has solid state hardware. So to me this was just embarrassing.

Really hope Drake can fix this issue.
decompression during loading is one thing, but games might trend towards loading large amount of assets on the fly, much like Ratchet and Clank Rift Apart. granted, such a thing has been done before, but not with such high fidelity assets. bespoke Drake games will get by, but ports of other games might struggle
 
decompression during loading is one thing, but games might trend towards loading large amount of assets on the fly, much like Ratchet and Clank Rift Apart. granted, such a thing has been done before, but not with such high fidelity assets. bespoke Drake games will get by, but ports of other games might struggle
I honestly don't think Drake will get ports that utilise that sort of tech, not if it is an absolute must as part of the experience and I imagine it will be.

Unless there is a floor that is set? Maybe 1GBPS read speed, Drake may be able to handle that just about if the asset quality takes a hit.

Will be interesting to see. I just don't want to have to wait over 30 seconds for a door to load in resident evil, only to then leave that room 6 seconds later haha.
 
0
Thanks for this, it was really informative. Nintendo could use RTXIO for shrinking loading times I imagine as the GPU shouldn't be doing too much at that time. This is really positive and this enhanced asset decompression speed could make faster storage such as UFS 3 more likely.
As @RennanNT points out, RTXIO is really two parts - one, GPU accelerated decompression, and two the ability to stream data straight to the GPU without having to go through main memory. The existing Switch can already do that second one. GPU decompression is a technique that, in theory, can be implemented in any hardware if the developer really wants to. The question is how easy will Nvidia/Nintendo make it.

I bring this up partially because, as a systems/server side technologist, this was something I misunderstood about console hardware myself. Direct Memory Access to the GPU by random software is generally a Bad Thing from a security perspective, and only really important for a few applications like games. In the case of the Switch which is a closed ecosystem (no random downloads from the internet) and only plays games, the priorities are different.
 
0
I'm not expecting 2022 at all and will think simultaneous release with botw 2 is a bit more likely, if not too predictable. But of course, anything goes.

With Switch not slowing down significantly in sales though, why would Nintendo release it this Fall opposed to Spring of the following year? What would they benefit from a Q4 release?. I can list a few reasons why from Nintendo's perspective why it could be better to push back.

But if Bayonetta 3 releases this year, that and Scarlett/violet are the perfect games to showcase them.

if Nintendo does go for a March-Spring 2023 release. Perhaps after Scarlet/Violet get released is the most strategic.. 🤔
They waited too long to launch the 3DS and Wii U granted it wasn't all greed and may have legitimately been hardware and getting games ready, but to avoid peaks and valleys (which iirc they have stated they wish to avoid) in their earnings they need to launch before Switch seriously declines.

With timing of the launch, my view is that Past practice isn't a good indicator and their stated financial goals may be more instructive In this case.
 
Switch OLED was announced in July, but at this time last year we already had manufacturing updates from Bloomberg stating that it had gone into production. So we knew new hardware was coming.

Right now it's just complete silence.
 
Switch OLED was announced in July, but at this time last year we already had manufacturing updates from Bloomberg stating that it had gone into production. So we knew new hardware was coming.

Right now it's just complete silence.
Silence from Bloomberg, but we did get a potential factory leak in April extremely similar to that of the Switch Lite, same time frame too.
 
Idk, I think a 4-8 week window for a device like this is cutting it way too close. Maybe 8 weeks I guess, I can’t remember what the Lite was, but no way it’s only a month in advance. Leaks for days.

The Lite was revealed in July and launched 2 months later.

The New 3ds was revealed in August and launched 6 weeks later.

I didn’t suggest it would me only a month. 6 weeks - 8 weeks is certainly extremely possible
 
0
I know, that’s why I mentioned 1 cluster along with the other cluster option. :p

It could be a config that is similar to one of the snapdragons that have 4X1+4A78, but in the case of Drake it could be 4A78+4A78

Could be 4 faster A78 cores and 4 slower A78 cores too 🤷🏾‍♂️

This wouldn’t be the A78C form. But the A78 form.

Does any smartphone, tablet or laptop use the A78C?
 
Been such a boring year and I currently feel very antsy wanting something exciting to happen like really badly. We're in the middle of a thunderstorm currently, I'm off of work for a couple days over the weekend and I'm in the mood for some exciting Nintendo news to drop or leak. I really hope June is the month things start getting more exciting for gaming altogether.
 
Can’t really shake this comment from Mochizuki’s September 2021 article:

Developers declined to speculate on Nintendo’s plans for another console but said they expect to release their 4K Switch games during or after the second half of next year.

Honestly it reads like it wasn’t that long ago that some developers confirmed their expectations of H2 this year. Regardless of Nintendo’s own software plans, this still feels like they’d be on the hook to deliver outside of some extraordinary circumstances.

We still have a very quiet H2 for Switch from both first and third parties (I think). Not much has changed since whatever they announced last year.

my view is that Past practice isn't a good indicator and their stated financial goals may be more instructive In this case.

Imran’s commented more than once similarly citing that leadership isn’t the same:

 
I know, that’s why I mentioned 1 cluster along with the other cluster option. :p

It could be a config that is similar to one of the snapdragons that have 4X1+4A78, but in the case of Drake it could be 4A78+4A78

Could be 4 faster A78 cores and 4 slower A78 cores too 🤷🏾‍♂️

This wouldn’t be the A78C form. But the A78 form.
Nintendo and Nvidia could also theoretically opt to use two voltage regulators for the Cortex-A78C, with one power rail dedicated to four performance (high frequency) CPU cores, and another power rail dedicated to four efficiency (low frequency) CPU cores, except only one CPU cluster is needed.

Two voltage regulators are necessary regardless of which Nintendo and Nvidia decide to use for the CPU (two Cortex-A78 clusters or one Cortex-A78C cluster) since that's how DynamIQ works.

Does any smartphone, tablet or laptop use the A78C?
No.

But Nvidia does seem to use the Cortex-A78C for the BlueField-3 since Nvidia mentions that the BlueField-3 has 16 MB for the system cache, which does imply that Nvidia uses two Cortex-A78C clusters, since the Cortex-A78C allows up to 8 MB of L3 cache.
 
0
Since the CPU talk has died down… 👇
Snice talk about Drake’s potential “gimmick” has come up, can we discuss the possibility of a camera(s)? AR utilization would need it, which Nintendo has dabbled with in the past. It could provide “a new way to play” that couldn’t be done on Switch before. And if Nintendo ever got into 3rd party apps again (Im still bitter about no Netflix on Switch), then Zoom or other camera based apps/games could be a possibility. Yes, that would dictate certain games/apps be handheld only, but I think that’s a fair trade. After the DSi, 3DS, and WiiU, I thought cameras were going to be a staple on all hardware going forward. But Switch bucked the trend. Has the price of decent mobile cameras decreased enough to make this a good business decision? IMO 480p cameras couldn’t cut it in 2011, but Nintendo went with them anyway. I wonder what resolution they would go with today.

EDIT

Depending on dock design, I suppose a front camera could be used docked as well. Maybe like a PSEye or Kinect scenario…
I can’t help but wonder if some of this extra power drake has is going to be utilized on something besides graphics (maybe a camera). Like how the new 3DS used extra power for face tracking.
 
Since the CPU talk has died down… 👇

I can’t help but wonder if some of this extra power drake has is going to be utilized on something besides graphics (maybe a camera). Like how the new 3DS used extra power for face tracking.
there's dedicated hardware for using the camera these days, though that might be a waste of silicon. but there would be plenty of gpu and cpu power to make up for that
 
0
Honestly it reads like it wasn’t that long ago that some developers confirmed their expectations of H2 this year. Regardless of Nintendo’s own software plans, this still feels like they’d be on the hook to deliver outside of some extraordinary circumstances.
My Mostly Evidence-less Theory: Nintendo distributes devkits, puts devs on hook contractually to deliver by hardware launch. OLED model surprises devs, subsequent behind the scenes chatter makes devs wonder if Nintendo is going to delay hardware. Mochizuki is poking around, frustrated devs talk.

Whatever happened, Nintendo has clearly put their foot down. If devs were told of a delay, they haven't been telling journalists (at least not in numbers that allow those journalists to run with the story). If devs haven't heard anything new, then even if they have been chatty, it doesn't give journalists juice behind a follow up story. Some folk have used the lack of rumor to disprove that things are imminent. I'm not convinced.

If you look at the pattern of rumor around the lite and the OLED there was a lot of rumor that something was coming, some of which were correct, some which were not, and some which were mixups (the OLED/Pro mashup). There were a few manufacturing leaks ~3 months prior to launch, but they were hard to discover. The famous one of the lite's backplate was caught by a western Twitter user, who spied it on a Chinese social media site, and didn't get picked up till after the Lite released and folk saw that it was real. The OLED model leaked to Mochizuki from manufacturers, and I don't recall any other details cropping up.

I'm not 100% convinced that this thing launches this year. I am 100% convinced that launching by the end of the fiscal (March 2023) is Nintendo's plan, and the absence of evidence isn't evidence of absence.
 
0
Switch Lite's case leak was in April, so 3 months before announcement, 5 months before release

I doubt the Drake form factor is any different than the OLED.

Any examples of picture leaks that don’t involve new form factor cases?

I’d be very surprised if any leaks from that time were still accessible.

I figured as much for that.

So still, nothing tells me that releasing a new model this Oct/Nov and revealing it in Sept is out of the question. Very possible, even.
 
0
Can’t really shake this comment from Mochizuki’s September 2021 article:



Honestly it reads like it wasn’t that long ago that some developers confirmed their expectations of H2 this year. Regardless of Nintendo’s own software plans, this still feels like they’d be on the hook to deliver outside of some extraordinary circumstances.

We still have a very quiet H2 for Switch from both first and third parties (I think). Not much has changed since whatever they announced last year.



Imran’s commented more than once similarly citing that leadership isn’t the same:


That chart is also looking at Japan, which is ironic as it is pretty saturated and sales had already dropped behind FY 2021 peaks last year. We know their forecast for FY2023 is down again from FY2022 , maybe not by as much as people expected but then again they wanted to sell 25 million (probably more in an unstated stretch goal) in 2022 and only managed 23~ million.
So some unfilled demand got shifted to this year, but I think once it starts dropping in the increments people expect, the new hardware should be announced and about ready to launch if not already launched. If it isn';t you're going to have 12-18 months transition period where your old hardware is dying/gone and your new one isn't out yet. Lost mineshare etc. Exactly the things with the DS and Wii transitions.

Not to mention how poorly it looks on their financials, basically a rollcoaster going straight down.
 
Last edited:
What actually happened to the "We should get some leaks after GDC?". There has been absolutely nothing after GDC wich was like 2 month ago? There is too little smoke for me right now anymore to be optimistic about a release this FY.

I agree there has been nothing heard about this thing since that nvidia leak, shows that this device is quite possibly not coming in 2022 or even considerably early 2023
 
0
What actually happened to the "We should get some leaks after GDC?". There has been absolutely nothing after GDC wich was like 2 month ago? There is too little smoke for me right now anymore to be optimistic about a release this FY.
Yeah its been super quiet. Nothing new from any insiders. Nothing new from all these media outlets.
 
0
Please read this staff post before posting.

Furthermore, according to this follow-up post, all off-topic chat will be moderated.
Last edited:


Back
Top Bottom