• 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.

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

yes. all that cost is just part of developing a chip. if they chose a different node at the start then they don't have to worry about redesigning the chip in the future

Isn’t that enough to assume it wouldn’t be the choice? What advantages could come with that that would outweigh the disadvantages?
 
Isn’t that enough to assume it wouldn’t be the choice? What advantages could come with that that would outweigh the disadvantages?
They’re not designing a chip from scratch, is the key piece of info.

Ampere, the GPU arch, is on 8nm. If they want to go to a different node Nintendo has to fully fund a port of NVidia’s last gen GPU arch to a new node, of which they would be the only customer.

Orin is on 8nm, which means not only would Nintendo need to fund the Ampere port, they would need to fund the entire chip design.

The costs that folks are (rightly) pointing out would have to be paid for a nodeshrunk Drake in the future would still have to be paid, they’d just be paid now. For a future node shrink that Nintendo might not need or want.
 
I mean we got the Linux update a few weeks back and now there’s this potentially interesting firmware update, so things have already been lining up.

7 months before the OG Switch launch would be about late July/ early August. And what happened around that time? Eurogamer’s NX blowout article was published. But obviously Philips would have received that inside info at least a few weeks before the article.

That would line up with the current timeframe of the Linux update and OS update being pushed just before 7 months ahead of launch. I assume similar things happened with the original Switch back in July of 2016 or so. Hardware was likely finalized and 1.0.0 was flashed onto launch units.

Is it just a coincidence? I don’t know, but this is my smoking gun. Probably. Maybe.
 
Last edited:
The GPU arch isn’t only on 8nm. Only Orin is.
But only the A100 yes? I’ve mentioned that I though TSMC 7nm the most likely candidate after Samsung 8nm for that reason, but the A100 isn’t desktop Ampere. It doesn’t have the same NVENC/DEC, it doesn’t have RT, it doesn’t support any graphics APIs. It just does CUDA/Tensor ops.
 
In everyone who has kept up with this' opinion, how far are we from an initial official announcement/acknowledgement?
General consensus seems to be release in early 2023, likely by May 12 at the latest, or if not then all bets are off and nobody knows.

Announcement probably early 2023, like Jan or Feb.
 
General consensus seems to be release in early 2023, likely by May 12 at the latest, or if not then all bets are off and nobody knows.

Announcement probably early 2023, like Jan or Feb.
The release window seems to be in line with what I'd expect, I just figured we'd be getting close to an announcement now. But hey, I guess maybe they don't want to undermine Holiday season sales.
 
0
It uses both, yes? My understanding was that Samsung's early EUV capacity was too low to build entire chips with, and that the majority of layers were still DUV. This is not my wheelhouse, so perhaps that is common knowledge or I am mistaken.
Yes, EUV lithography wasn't used for all the layers for Samsung's 7LPP process node. Specifically, the fin in Samsung's 7LPP process node still uses DUV lithography (ArF). But Samsung still mentions that Samsung's 7LPP process node uses EUV lithography.

And EUV lithography wasn't used for all the layers for TSMC's N6 process node as well. But TSMC still mentions that TSMC's N6 process node uses EUV lithography.
 
But only the A100 yes? I’ve mentioned that I though TSMC 7nm the most likely candidate after Samsung 8nm for that reason, but the A100 isn’t desktop Ampere. It doesn’t have the same NVENC/DEC, it doesn’t have RT, it doesn’t support any graphics APIs. It just does CUDA/Tensor ops.
It matters more that the Nvidia engineers exist at the fab than that, as they are the ones who implemented the IP at TSMC. If they didn’t exist at said fab this would be tricky to work with. They’d understand how to implement almost related to Ampere IP within the TSMC 7nm IP.


For tangentially related, Qualcomm had the SnapDragon 8 Gen 1 at Samsung 4nm, and within a year or so they had 8+ Gen 1 at TSMC. Now, someone would ask, “how did they migrate it in such a short time?” the answer is that they didn’t. They had two engineering teams working at two different foundries and aiming to build the chip with the engineers of the company and the engineers of the foundry and the ARM engineers.

Qualcomm is historically pretty much like nVidia in being a fabless entity that necessitates being flexible in situations like this.
In everyone who has kept up with this' opinion, how far are we from an initial official announcement/acknowledgement?
I’m not really sure anymore. It’s up in the air at this point.

It’ll be shown whenever they feel like it I feel.

It can be an “any day now” thing or “in the next 18-24 months”
 
Last edited:
In everyone who has kept up with this' opinion, how far are we from an initial official announcement/acknowledgement?
Anytime after the launch of FEE and the end of their fiscal year. Earliest being February and the latest being March.

Also, since the SWOLED had a three month announcement to launch gap, I think that is the minimum for Drake. So the latest I see this being fully revealed is mid February.
 
Last edited:
In everyone who has kept up with this' opinion, how far are we from an initial official announcement/acknowledgement?
Within the next 120 days. When in those next 120 days is anyone's guess, but that's the time window we should be looking at.
 
Regarding Samsung 7nm (non-EUV) yesterday, I found that he made the same post in March.
He is very familiar with Samsung. So it is not natural to say something that does not exist...
 
Regarding Samsung 7nm (non-EUV) yesterday, I found that he made the same post in March.
He is very familiar with Samsung. So it is not natural to say something that does not exist...
Does “familiar” imply credibility?
 
Regarding Samsung 7nm (non-EUV) yesterday, I found that he made the same post in March.
He is very familiar with Samsung. So it is not natural to say something that does not exist...
I’d like to stress it’s very possible this is referring to the 8nm node.

And that has been mistakenly reported as 7nm in the past by foreign outlets.



There is one other possibility but that depends on how people would believe this as possible: nvidia worked with Samsung to further enhance the 8N node to be more like the 7nm in performance and power savings. So, just like how nvidia customized a 10nm for their own needs in 8N, 7N is a custom Nvidia only node. But this would mean stretching imagination a bit to conform to this possibility. Then again, as @Dakhil reminds us, node nomenclature is simply a marketing term. So, this “7nm” theory could really just be 8nm anyway. There’s not really anything to further go along with this though as it’s just further refining the 10nm process beyond 8N, I think the next possible option would be Samsung’s 3nm family for a potential die shrink if it remains with SEC.

Something like how the Wii U is a further refined GameCube in some aspects, not all, some. That’s how it could be for 8N to this theoretical “7N” node.
 
0
Anytime after the launch of FEE and the end of their fiscal year. Earliest being February and the latest being March.

Also, since the SWOLED had a three month announcement to launch gap, I think that is the minimum for Drake. So the latest I see this being fully revealed is mid February.
I understand not wanting to complicate things for holiday sales, but trying to aim for a several month gap without much first-party base Switch software going on just seems like... a bad software lineup.
 
I understand not wanting to complicate things for holiday sales, but trying to aim for a several month gap without much first-party base Switch software going on just seems like... a bad software lineup.
Some posts already argued on this but I really don’t think a 2022 reveal would lower holiday sales that much, especially with a new Pokémon generation in November.
And most of the loss would technically be delayed purchases for Drake in 2023 anyway.
It’s the fiscal year that counts mostly, isn’t it?
 
Some posts already argued on this but I really don’t think a 2022 reveal would lower holiday sales that much, especially with a new Pokémon generation in November.
There is huge sales evidence to the contrary.

And most of the loss would technically be delayed purchases for Drake in 2023 anyway.
It’s the fiscal year that counts mostly, isn’t it?
If it's the fiscal year that counts, and you're releasing hardware after March, then you're tanking holiday sales, with no surge at the end of the fiscal. And even if that is the right call, you get to be yelled at by investors about what tank holiday sales when Switch is still going well?

I don't know what Nintendo's strategy is for this thing, but I don't see any upside to announcing this year.
 
There is huge sales evidence to the contrary.


If it's the fiscal year that counts, and you're releasing hardware after March, then you're tanking holiday sales, with no surge at the end of the fiscal. And even if that is the right call, you get to be yelled at by investors about what tank holiday sales when Switch is still going well?

I don't know what Nintendo's strategy is for this thing, but I don't see any upside to announcing this year.
Yeah, I see. Of course I was implying a Drake release before the end of March to compensante but no way to be sure about that :(
 
Yeah, I see. Of course I was implying a Drake release before the end of March to compensante but no way to be sure about that :(
I think October Reveal, March Release makes sense. I don't see a November or December reveal being likely, so if it's this year I'd say this month.

After that, January to March reveal for a March to May release, most likely.

But I wouldn't believe anything about "tanking holiday sales", that's pretty much ahistoric, perhaps even apocryphal. 6 or 7 months of build up won't Osbourne Effect the Switch out of the market, trust me.

As I've gone into detail about before:
If they EVER planned to possibly release this in 2022, if won't miss the Fiscal Year. Missing quarterly launch targets is reasonable, missing the entire fiscal year is unprecedented, for hardware at least.
 
It's not gonna happen this year (the announcement, that is).
Like @oldpuck said already, there's evidence (eg: vgcharts) on how announcements/releases of next gen consoles affected sales of the previous gen.
This was the case with the DS -> 3DS, new 3DS -> switch. And before someone mentions how "the 3ds kept selling even after the switch's release" I honestly ask you to look at how 3DS sales tanked ~1 year (more or less) after the switch's release.
And if you're thinking about the whole thing they had going with special ed. 3ds consoles released just a few months before the new 3ds reveal (which is kind of similar to the current situation we are in with the switch), I say don't worry too much about it.
That scenario isn't indicative of anything since back then I wasn't sure, but today I'm pretty confident that with the new 3ds, nintendo just wanted to make a definitive version of the console, not an iteration that brought extensive software exclusivity or even a "next gen" device lineup.
Which looking at the OLED switch is... well, kind of the same idea if you think about it really.

January~March reveal with a launch in May~July is my bet.
 
If Drake isn't going to be called a straight Switch 2 and sold as a new generation. I dont see why they need a long period from announcement to launch. 2 months would be ample time. What do we need 6 months gap for?
 
But I wouldn't believe anything about "tanking holiday sales", that's pretty much ahistoric, perhaps even apocryphal.

On what basis are you making this statement? If you compare across years for the same quarter, and account for downward trends overall - the fairest way to account for this effect on the market - the Switch drove a 20% drop in 3DS sales, the 3DS drove a 50% drop in DS sales,

6 or 7 months of build up won't Osbourne Effect the Switch out of the market, trust me.
No, it won't. Sales tend to restore to normal in the following quarter if there are multiple quarters between announcement and launch of the followup, but it does leave a net impact on yearly sales.

There are perfectly good reasons to bite that bullet, but that's not to say it doesn't exist.
 
0
The thing with Nintendo consoles is that we have (essentially) 3 kinds of consumers:
  • The people who want to buy the console for themselves and are either financially independent or, managed to save enough money.
  • The parents who are looking for something to buy to their children.
  • The ones who are looking to buy something as a gift (be it a family member or a friend)

The first group is imo, the most economically important according to that one data released by nintendo on their age group research. I also think so considering the Nintendo fanbase is an aging group by now with many financially independent people on their mid 20's-early 30's.
The second group and the third is where the key reason as to why I think that @Concernt 's argument might be a solid one and here's why:
Those two kinds of consumers are the ones who are looking for an immediate option, if black friday, an anniversary or Christmas hits, they won't wait until next year for an announcement/release of a product no one knows to make a purchase.

If a switch pro were announced TODAY but, only released next year groups #2 and #3 would still (mostly) all just buy the OLED/v2 or Lite models anyways.
However, I won't be convinced of a 2022 announcement/release until someone hands me proof that groups #2 and #3 account for over 50% of switch purchases.

And I honestly think that Nintendo has such kind of data and will make their decision based on (not only) that.

edit: another counterargument is that group #3 is price-focused as it's not that common for a friend/non-core family member to drop 300$+ on a gift. If the switch pro releases at an even higher price range than the OLED, you can be sure that group #3 will mostly be out of the equation.
 
Last edited:
If Drake really uses 8nm, then can we safely assume it’s going to be a “Pro” instead of a “2”? It sounds like in order to keep battery life and heat in check, the clocks would have to be pretty low on 8nm, meaning less TFLOPs. I forgot where all the calculations are for TFLOPs on various nodes, but I guess my concern is: can Drake, with all the details we know, still be considered “next gen” if it’s limited to 8nm?
 
If Drake really uses 8nm, then can we safely assume it’s going to be a “Pro” instead of a “2”? It sounds like in order to keep battery life and heat in check, the clocks would have to be pretty low on 8nm, meaning less TFLOPs. I forgot where all the calculations are for TFLOPs on various nodes, but I guess my concern is: can Drake, with all the details we know, still be considered “next gen” if it’s limited to 8nm?
No.

The clocks on 8nm won't be all that different from clocks on 5nm, there is a lower bound they can't really pass and an upper bound that comes up pretty quickly regardless of node. We already more or less have the full specs, it gives us a pretty narrow range of performance, this thing will not be able to be all that different just based on clocks. If they want to call it a "pro" or a "2" or a ganache is totally up to them, the process node is like, the least likely determining factor in that decision.

As I said yesterday the worry over node is getting a bit silly.
 
No.

The clocks on 8nm won't be all that different from clocks on 5nm, there is a lower bound they can't really pass and an upper bound that comes up pretty quickly regardless of node. We already more or less have the full specs, it gives us a pretty narrow range of performance, this thing will not be able to be all that different just based on clocks. If they want to call it a "pro" or a "2" or a ganache is totally up to them, the process node is like, the least likely determining factor in that decision.

As I said yesterday the worry over node is getting a bit silly.
That’s not true. There is a huge difference between 5 nm and 8. Depending on the estimate you are looking at, 8 nm would struggle to run 12 sm at all in portable mode.
 
If Drake really uses 8nm, then can we safely assume it’s going to be a “Pro” instead of a “2”? It sounds like in order to keep battery life and heat in check, the clocks would have to be pretty low on 8nm, meaning less TFLOPs. I forgot where all the calculations are for TFLOPs on various nodes, but I guess my concern is: can Drake, with all the details we know, still be considered “next gen” if it’s limited to 8nm?
Pro/ 2 doesn’t have all that much to do with power, and a lot more to do with marketing.

Edit: and even at 8nm, with sm disabled in portable mode there’s no way it won’t run circles around the current Switch.
 
Last edited:
As much as we'd like confirmation in this speculation thread, there's no upside to publicly announcing a Switch successor/Pro earlier than necessary to avoid significant supply chain/shipping leaks, and certainly not before Christmas time sales. An announcement in February or March is reasonable for a May release.
 
0
That’s not true. There is a huge difference between 5 nm and 8. Depending on the estimate you are looking at, 8 nm would struggle to run 12 sm at all in portable mode.
There's a huge difference between different products made on those nodes. If you're talking about the same chip theoretically being made twice once on each process the comparison between the two wouldn't be so far apart. Yes, it will be more power efficient. Not double or anything drastic.

But what exactly do you get from more power efficiency? Power does not scale linearly with clock speed, you can only increase it so much before your power consumption reaches the maximum possible for the form factor/battery.


All I'm trying to say here is that the actual physical configuration of the chip tells us much more about what it will be capable of than the process its made on.
 
There's a huge difference between different products made on those nodes. If you're talking about the same chip theoretically being made twice once on each process the comparison between the two wouldn't be so far apart. Yes, it will be more power efficient. Not double or anything drastic.

But what exactly do you get from more power efficiency? Power does not scale linearly with clock speed, you can only increase it so much before your power consumption reaches the maximum possible for the form factor/battery.


All I'm trying to say here is that the actual physical configuration of the chip tells us much more about what it will be capable of than the process it’s made on.
I don’t think double is that far from the truth actually.
 
0
Everyone knows they are going to announce it next week.
RvG.gif
 
This device is a little overdue for my personal tastes. Tunic, No Man's Sky and Prodeus are three ports I'd love to play at a better render quality than what's currently possible on Switch. Don't wanna game on a different platform thank you very much, as Switch is simply the user friendliest gaming device I know, but Doom 2016 style blur at 30FPS is getting real old at this point =P . Splatoon 3 felt like a great shot in the arm subjectively, as it's a beautiful and sharp game at 60FPS, but it's the exception.
 
We all know Nintendo can market it however they want, and they can squeeze every ounce of power to make their first party games look amazing, but I’m wondering about the possibility of “miracle” PS5/XSX ports on 8nm Drake. Ah well. I’ll just wait and see. My last thought is if 8nm Drake gets the “Mariko“ treatment and gets a die shrink later on, it won’t be for the benefit of more power but to battery life again. Versus if they did 6nm now, then 3nm in a few years.
 
It's not gonna happen this year (the announcement, that is).
Like @oldpuck said already, there's evidence (eg: vgcharts) on how announcements/releases of next gen consoles affected sales of the previous gen.
This was the case with the DS -> 3DS, new 3DS -> switch. And before someone mentions how "the 3ds kept selling even after the switch's release" I honestly ask you to look at how 3DS sales tanked ~1 year (more or less) after the switch's release.
And if you're thinking about the whole thing they had going with special ed. 3ds consoles released just a few months before the new 3ds reveal (which is kind of similar to the current situation we are in with the switch), I say don't worry too much about it.
That scenario isn't indicative of anything since back then I wasn't sure, but today I'm pretty confident that with the new 3ds, nintendo just wanted to make a definitive version of the console, not an iteration that brought extensive software exclusivity or even a "next gen" device lineup.
Which looking at the OLED switch is... well, kind of the same idea if you think about it really.

January~March reveal with a launch in May~July is my bet.
May ~ July is in no realm "early 2023". If we are to believe the leaks we see of Drake, and they all point to early 2023, what sense does it make to suggest May ~ July?

And an announcement really wouldn't affect sales that much. 3DS as an example had an amazing 2016.
 
And an announcement really wouldn't affect sales that much. 3DS as an example had an amazing 2016.
It affects sales for the group of people who are interested in either buying the switch v2 or OLED this black friday or christmas.
Which is quite a lot of people considering how every BF/xmas the sales for switch consoles skyrocket compared to other months.
May ~ July is in no realm "early 2023". If we are to believe the leaks we see of Drake, and they all point to early 2023, what sense does it make to suggest May ~ July?
Yeah, I get that by then it's mid 2022. It's just that I'm really used to my uni's terminology per 2 semesters - early 2022/late 2022.
 
0
Hi there!

Are there any rumors/news about the power consumption or the new cooling techs of the new hardware?

The current switch power consumption is approx 6w when playing games on docked mode. If the volume of the new console almost the same size with the current one, and if there is no significant upgrade of the heat cooling, then it seems that the new console would be also in the same level of power consumption.
I'm a little bit worry about the performance at this low tdp.
 
I guess my concern is: can Drake, with all the details we know, still be considered “next gen” if it’s limited to 8nm?
I'm a little bit worry about the performance at this low tdp.
Depending on how the nanometer stuff pans out, at the extreme ends we might be looking at "Ridiculously more powerful than 2017 Switch" and "Ridiculously more powerful than 2017 Switch times 1.5".
 
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