• 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!)

God these USB names are so dumb… but yes, USB 3.2 Gen 2. Though it could be USB 3.2 Gen 2x2 which offers more bandwidth if they felt like they needed it. 2.5GB/s theoretically speaking.




I was gonna ask about these, but since the switch will most likely not utilize the 120Hz feature, would it even matter if it’s at 60Hz to get the full 4:4:4 and full bit color depth for HDR? Wouldn’t this be a scenario in which them not going with the full usage of 120Hz, even if they support it, be a benefit as they have some bandwidth to spare?


Fake edit: seems like even if Nintendo were to support “4k” at 60Gz for all their content (at the highest) and to get 12-bit 4:4:4 even at 60Hz, they’ll need to support at least HDMI 2.1. Otherwise they’ll be limited to 4:2:0 chroma.


I wonder if they lowered the resolution, would it be enough… with DLSS not actually being that resolution if it has any effect on this.
I'm almost certain it'll be HDMI 2.0, not least because Switch is internally DisplayPort so they'd need to engineer a whole new crossbar setup that might interfere with dock compatibility between the different models. Plus as I've said before I fully expect them to use the Nintendo Switch Dock with LAN Port rather than a new "Nintendo Switch Dock 4K", when the existing Dock (with LAN Port) already supports HDMI 2.0. This device can't really be any thicker than the current Switch without breaking Joy-Con compatibility, will definitely use USB C again, etc.

There's no reason for them to move to a new Dock so soon after the last one when reusing the black Dock with LAN Port reduces the number of unique parts, reduces costs, had so much engineering gone into it, has 4K output, and has more ventilation than the Nintendo Switch V1, V2 or OLED can even use. Why would they engineer the dock that comes with the OLED Model to have MORE ventilation when it's a more power efficient model?

Dock redesigns every 1-2 years just doesn't seem like the route they'll go when the dock they have is HDMI 2.0b and can do 4K60HDR and surround sound without breaking a sweat! I doubt the device will even attempt to target 120FPS.
 
I was gonna ask about these, but since the switch will most likely not utilize the 120Hz feature, would it even matter if it’s at 60Hz to get the full 4:4:4 and full bit color depth for HDR? Wouldn’t this be a scenario in which them not going with the full usage of 120Hz, even if they support it, be a benefit as they have some bandwidth to spare?


Fake edit: seems like even if Nintendo were to support “4k” at 60Gz for all their content (at the highest) and to get 12-bit 4:4:4 even at 60Hz, they’ll need to support at least HDMI 2.1. Otherwise they’ll be limited to 4:2:0 chroma.


I wonder if they lowered the resolution, would it be enough… with DLSS not actually being that resolution if it has any effect on this.
well, 4K out put is 4K output, regardless of how one gets there. but Nintendo can always go in between 1440p and 2160p and have 4:4:4 if the bandwidth allows it. and just let the tv get the rest of the way. it's not like anyone but the most tech whorey can see the difference
 
Hi folks! Are there any new relevant findings since last month? Pretty busy lately couldn’t keep up with the thread.
The last one was an unverified factory uncle in Korea saying it’s Samsung 7nm, but not EUV. Since there is no such thing, it could just mean 8nm. Also it could be BS.
 
Last edited:
I'm almost certain it'll be HDMI 2.0, not least because Switch is internally DisplayPort so they'd need to engineer a whole new crossbar setup that might interfere with dock compatibility between the different models. Plus as I've said before I fully expect them to use the Nintendo Switch Dock with LAN Port rather than a new "Nintendo Switch Dock 4K", when the existing Dock (with LAN Port) already supports HDMI 2.0. This device can't really be any thicker than the current Switch without breaking Joy-Con compatibility, will definitely use USB C again, etc.

There's no reason for them to move to a new Dock so soon after the last one when reusing the black Dock with LAN Port reduces the number of unique parts, reduces costs, had so much engineering gone into it, has 4K output, and has more ventilation than the Nintendo Switch V1, V2 or OLED can even use. Why would they engineer the dock that comes with the OLED Model to have MORE ventilation when it's a more power efficient model?

Dock redesigns every 1-2 years just doesn't seem like the route they'll go when the dock they have is HDMI 2.0b and can do 4K60HDR and surround sound without breaking a sweat! I doubt the device will even attempt to target 120FPS.
The dock only handles the DisplayPort (coming from the console) to HDMI signal conversion. The USB signal to DisplayPort signal switching is handled by the console.

So if VRR is an important enough feature that necessitates Nintendo to use HDMI 2.1, especially since HDMI 2.0b doesn't natively support VRR, unless FreeSync is used, then Nintendo only needs to replace the crossbar switch chip on the console (e.g. PI3USB30532 [USB 3.0/DisplayPort 1.2] to PI3USB31532 [USB 3.2 Gen 2/DisplayPort 1.4]) and the DisplayPort to HDMI converter chip on the dock (e.g. RTD2172N [DisplayPort 1.4 to HDMI 2.0b] to RTD2173 [DisplayPort 1.4 to HDMI 2.1]).

The OLED model still uses the PI3USB30532 chip, which has been used on the Nintendo Switch. However, the HDMI 2.0 cable provided by Nintendo for the OLED model's dock is apparently causing blackouts, which seems to have been fixed by switching to a HDMI 1.4 cable, which makes me wonder if the PI3USB30532 chip is the culprit for the blackouts. If so, then Nintendo has to change the crossbar switch chip on Nintendo's new console regardless.

And VRR doesn't necessarily require 120 Hz support. The only reason 120 Hz is mentioned is that all the mobile OLED displays with VRR support happen to support 120 Hz (e.g. iPhone 13 Pro and iPhone 13 Pro Max).
 
The OLED model still uses the PI3USB30532 chip, which has been used on the Nintendo Switch. However, the HDMI 2.0 cable provided by Nintendo for the OLED model's dock is apparently causing blackouts, which seems to have been fixed by switching to a HDMI 1.4 cable, which makes me wonder if the PI3USB30532 chip is the culprit for the blackouts. If so, then Nintendo has to change the crossbar switch chip on Nintendo's new console regardless.
Is there nothing to suggest that a software update could be sent out to fix this blackout issue rather than fully updating this chip? Truthfully it's the first I heard of this problem; I don't doubt it exists, I just didn't realize it was so widespread.
 
The last one was an unverified factory uncle in Korea saying it’s Samsung 7nm, but not UAV. Since there is no such thing, it could just mean 8nm. Also it could be BS.
Minor correction, but it's EUV (extreme ultraviolet) which they're claiming is not used.
 
well, 4K out put is 4K output, regardless of how one gets there. but Nintendo can always go in between 1440p and 2160p and have 4:4:4 if the bandwidth allows it. and just let the tv get the rest of the way. it's not like anyone but the most tech whorey can see the difference
TVs can be really picky about resolution. Not being exactly 4k would cause compatibility issues.
 
We could reeeeeally use some insider info right now...
On a side note: according to vgchartz, the switch is nearing 114mi units sold. November and December are peak months with iirc, more than double the average sold each month. I can totally see it surpassing the PS4 in mid~late December.
MERRY CHRISTMAS 🎄
 
0
We could reeeeeally use some insider info right now...
On a side note: according to vgchartz, the switch is nearing 114mi units sold. November and December are peak months with iirc, more than double the average sold each month. I can totally see it surpassing the PS4 in mid~late December.
If it doesn't pass the ps4 this year, it means something went very wrong. Official numbers will be out in 3 weeks anyway.
 
i think in the next 2/3 months we should really start hearing news about next HW if its still on track to release H1 2023.

Especially news from whatever games are being ported exclusive to next HW. This is honestly what most gamers with not so many HW tech knowledge wants to know.
 
We could reeeeeally use some insider info right now...
On a side note: according to vgchartz, the switch is nearing 114mi units sold. November and December are peak months with iirc, more than double the average sold each month. I can totally see it surpassing the PS4 in mid~late December.
The Switch was over 111 mi as of end of June. The PS4 latest number is 117 mi and it's not going up much. That's a 6 million difference.

Switch's worse Jul-Dec period was over 10 mi (7.2 mi of it from holidays quarter alone), which was in it's launch year when it was suffering hard from shortages. So, it's extremely unlikely it will not outsell the PS4 by the end of the year. And it will almost likely happen sometime in November, not December, though I'm not really expecting Nintendo to release a PR bragging about it, so we won't know when it happens exactly.
 
Is there nothing to suggest that a software update could be sent out to fix this blackout issue rather than fully updating this chip? Truthfully it's the first I heard of this problem; I don't doubt it exists, I just didn't realize it was so widespread.
I don't know. But if a software update doesn't fix the black out problem, and the culprit is indeed the PI3USB30532 chip, then Nintendo has no choice, but to update the crossbar switch chip.



This is not a chip we designed entirely from scratch. Rather, it's the scaled version of an already proven AI accelerator built into our Telum chip. The 32 cores in the IBM AIU closely resemble the AI core embedded in the Telum chip that powers our latest IBM's z16 system. (Telum uses transistors that are 7 nm in size while our AIU will feature faster, even smaller 5 nm transistors.)
Sounds like IBM have confirmed transitioning from Samsung's 7HPP process node to Samsung's 5 nm** process node, which seems like more confirmation Samsung's 7 nm** process node's being discontinued.

** → a marketing nomenclature used by all foundry companies
 
is that a problem with modern tvs? that's what the upscalers are for, no? same for tvs that accept PCs as an input and can be fed all sorts of weird resolutions
My TV literally won't even accept a 1440p signal unless it's 120Hz, and it's a 2018 model. Honestly I don't think even it's intentional that it allows that through.
 
0
136 GB/s is 5.3x as much as lpddr4. Man I want this so bad on Drake... We need as much bandwidth we can get. Did they give a release date? I think lpddr5x is supposed to be out around q1 2023. Whatever hardware we are getting is likely finalized at this point though, and we likely barely missed the LPDDR5x window and have to settle with LPDDR5... sigh
 
Last edited:
0
On a side note: according to vgchartz, the switch is nearing 114mi units sold. November and December are peak months with iirc, more than double the average sold each month. I can totally see it surpassing the PS4 in mid~late December.
Definitely, if not earlier. By shipments, through June Switch is about 6 million short of the final reported PS4 number. Last year in July-December they shipped over 14 million.
 
That's cray. It'll be interesting to see the financial report after the holiday season.
Yeah they should be on around 120 million by the first '23 financial results. If this new device is "Switch: XXX" and not "Switch 2" then they have a real chance of surpassing PS2 by the end of 2025.
 
0
Alright, look alive people: tomorrow is Thursday, October 20. Nintendo Switch was revealed on Thursday, October 20 in 2016. Be prepared for "First look at Super Nintendo Switch" tomorrow morning.
 
Alright, look alive people: tomorrow is Thursday, October 20. Nintendo Switch was revealed on Thursday, October 20 in 2016. Be prepared for "First look at Super Nintendo Switch" tomorrow morning.
Can't wait for Miyamoto to appear and tell the audience "We here at Nintendo are always looking at the future... but there's one man who's looking at the future more than all of us. This future predictor of Nintendo products is now executive lead of our product development team, so without further ado, I present to you NateDrake, who predicted the Nintendo Drake Switch Console." Everybody will clap, and we'll find out NateDrake's true name is actually Nathan Drake.
 
Is it true that the T239 is a T234 but with a downgrade in its performance or power?
Not exactly. T239 has fewer CPU and GPU cores than T234 but it also has some other things IIRC and likely has most of the automotive components removed. They're not exactly the same chip, just one having lower performance.
 
Can't wait for Miyamoto to appear and tell the audience "We here at Nintendo are always looking at the future... but there's one man who's looking at the future more than all of us. This future predictor of Nintendo products is now executive lead of our product development team, so without further ado, I present to you NateDrake, who predicted the Nintendo Drake Switch Console." Everybody will clap, and we'll find out NateDrake's true name is actually Nathan Drake.
And then all of Famiboards cheers

And then Tom Holland walks into frame

And then all of Famiboards groans
 
DF tech review of Mario+Rabbids : Sparks of hope:


I can't help but feel like I am looking at a Switch port of a PS4 game(I am not talking about in comparison with the first game, even though Sparks of hope with its extended areas could have also looked like that on better hardware). But, like the game was intended to run on better hardware(but couldn't). Remember Drake could have been launched by now per Nate and Bloomberg and TOTK was first dated as a 2022 release.

I can't be the only one feeling that way...
 
DF tech review of Mario+Rabbids : Sparks of hope:


I can't help but feel like I am looking at a Switch port of a PS4 game(I am not talking about in comparison with the first game, even though Sparks of hope with its extended areas could have also looked like that on better hardware). But, like the game was intended to run on better hardware(but couldn't). Remember Drake could have been launched by now per Nate and Bloomberg and TOTK was first dated as a 2022 release.

I can't be the only one feeling that way...

Weren't there a few industry people making sly comments when this game was revealed about how it looked too good to be running on standard Switch hardware?
 
Remember Drake could have been launched by now per Nate and Bloomberg and TOTK was first dated as a 2022 release.
I believe NateDrake mentioned Nintendo targeting a launch window of holiday 2022 to early 2023 for Nintendo's new hardware. So I don't know if NateDrake necessarily implied Nintendo's new hardware could have launched by now, especially as NateDrake leaned more towards early 2023.
 
I believe NateDrake mentioned Nintendo targeting a launch window of holiday 2022 to early 2023 for Nintendo's new hardware. So I don't know if NateDrake necessarily implied Nintendo's new hardware could have launched by now, especially as NateDrake leaned more towards early 2023.
He did definitely imply it could have launched by now. He said Nintendo was targeting that as a best case scenario.

Also he kind of said H1 and early interchangeably.
 
Last edited:
DF tech review of Mario+Rabbids : Sparks of hope:


I can't help but feel like I am looking at a Switch port of a PS4 game(I am not talking about in comparison with the first game, even though Sparks of hope with its extended areas could have also looked like that on better hardware). But, like the game was intended to run on better hardware(but couldn't). Remember Drake could have been launched by now per Nate and Bloomberg and TOTK was first dated as a 2022 release.

I can't be the only one feeling that way...


Switch was always the target console
 
He did definitely imply it could have launched by now. He said Nintendo was targeting that as a best case scenario.

Also he kind of said H1 and early interchangeably, while it overlaps it isn’t the same.
He also said he was told that Sparks of Hope wouldn't make it out this year, so 🤷‍♂️

I don't think there is anything shown that implies it was designed for better hardware in mind though.
 
He also said he was told that Sparks of Hope wouldn't make it out this year, so 🤷‍♂️

I don't think there is anything shown that implies it was designed for better hardware in mind though.
I agree with that. Lower scale/ fixed perspective allowed them to make things prettier in the first one.
 
Kingdom Battle is arguably the best looking Switch game. I dunno why people leave it off their list. Sparks of Hope going for full camera control in a bigger RPG styled world required some sacrafices. That said it still looks like there is a bunch of environmental detail and geometry present. Side grade seems like the correct term. Would have been nice to see some reconstruction techniques though similar to XC3 or so TAA.

The game still looks great. Just couldn't pack in the same detail with the expanded scope. Bummer but I mean again, Kingdom Battle has imo always been the best looking Switch game. The amount of geometry in that game was crazy and it along with Luigi's Mansion 3 are the Switch games that look a lot closer to straight PS4/Xbox One games with little compromises.
 
So, it looks like Drake has a File Decompression Engine, a new piece of hardware which isn't on Orin.

A while ago after it was found out that Drake is using an 8 core CPU from a Linux commit message, I had a browse around myself and found some differences in some of the hardware blocks between Orin and Drake. Most of these are pretty understandable (removing automotive-focussed hardware), but one of them has stuck in my brain for a while, which is that there's a new block on Drake that's not on Orin, labelled FDE. I find this interesting because if it's on Drake but not on Orin, it would likely be something requested by Nintendo, or at least something that's of particular usefulness for a games console.

I just found a commit which solves the conundrum. The commit message simply reads:

Enable File Decompression Engine (FDE) driver as module.

I've also found a LinkedIn profile of a Nvidia employee which mentions FDE being "File Decompression Engine for games". I won't share it here, as I don't really feel comfortable posting links to random people's social media on public forums, but you can find it easily enough if you search. In any case, if the block is on Drake but not Orin, then it's not a huge leap to say it's for games.

A File Decompression Engine makes a lot of sense for a games console. Decompressing files as they're read from disk has a significant CPU overhead, something Nintendo are clearly well aware of, as Switch features a CPU boost mode specifically for improving loading times. It's a clear win for a fixed-function block like this, as modern game engines are constantly loading (and decompressing) assets from disk, and doing so on dedicated hardware will be more efficient in terms of transistors and power than throwing more CPU cores at it. Sony have gone this route with the PS5, where they've licensed both an algorithm and the hardware IP from an external source, whereas Nvidia seem to have done this in-house.

I think this is a pretty good sign for the next Switch. I've been saying for quite a while that dedicated decompression hardware would be a sensible thing to add to the new console, so it's good to see Nintendo and Nvidia thinking along the same lines. Of course it doesn't mean anything like PS5 load speeds, and we're still going to be limited by the read speeds of game cards, internal storage and removable storage, but it suggests Nintendo are serious about improving load speeds and decreasing CPU overhead of asset streaming. Which bodes well for them also using faster storage media too.
 
So, it looks like Drake has a File Decompression Engine, a new piece of hardware which isn't on Orin.

A while ago after it was found out that Drake is using an 8 core CPU from a Linux commit message, I had a browse around myself and found some differences in some of the hardware blocks between Orin and Drake. Most of these are pretty understandable (removing automotive-focussed hardware), but one of them has stuck in my brain for a while, which is that there's a new block on Drake that's not on Orin, labelled FDE. I find this interesting because if it's on Drake but not on Orin, it would likely be something requested by Nintendo, or at least something that's of particular usefulness for a games console.

I just found a commit which solves the conundrum. The commit message simply reads:



I've also found a LinkedIn profile of a Nvidia employee which mentions FDE being "File Decompression Engine for games". I won't share it here, as I don't really feel comfortable posting links to random people's social media on public forums, but you can find it easily enough if you search. In any case, if the block is on Drake but not Orin, then it's not a huge leap to say it's for games.

A File Decompression Engine makes a lot of sense for a games console. Decompressing files as they're read from disk has a significant CPU overhead, something Nintendo are clearly well aware of, as Switch features a CPU boost mode specifically for improving loading times. It's a clear win for a fixed-function block like this, as modern game engines are constantly loading (and decompressing) assets from disk, and doing so on dedicated hardware will be more efficient in terms of transistors and power than throwing more CPU cores at it. Sony have gone this route with the PS5, where they've licensed both an algorithm and the hardware IP from an external source, whereas Nvidia seem to have done this in-house.

I think this is a pretty good sign for the next Switch. I've been saying for quite a while that dedicated decompression hardware would be a sensible thing to add to the new console, so it's good to see Nintendo and Nvidia thinking along the same lines. Of course it doesn't mean anything like PS5 load speeds, and we're still going to be limited by the read speeds of game cards, internal storage and removable storage, but it suggests Nintendo are serious about improving load speeds and decreasing CPU overhead of asset streaming. Which bodes well for them also using faster storage media too.
this is a great find, very exciting
 
So, it looks like Drake has a File Decompression Engine, a new piece of hardware which isn't on Orin.

A while ago after it was found out that Drake is using an 8 core CPU from a Linux commit message, I had a browse around myself and found some differences in some of the hardware blocks between Orin and Drake. Most of these are pretty understandable (removing automotive-focussed hardware), but one of them has stuck in my brain for a while, which is that there's a new block on Drake that's not on Orin, labelled FDE. I find this interesting because if it's on Drake but not on Orin, it would likely be something requested by Nintendo, or at least something that's of particular usefulness for a games console.

I just found a commit which solves the conundrum. The commit message simply reads:



I've also found a LinkedIn profile of a Nvidia employee which mentions FDE being "File Decompression Engine for games". I won't share it here, as I don't really feel comfortable posting links to random people's social media on public forums, but you can find it easily enough if you search. In any case, if the block is on Drake but not Orin, then it's not a huge leap to say it's for games.

A File Decompression Engine makes a lot of sense for a games console. Decompressing files as they're read from disk has a significant CPU overhead, something Nintendo are clearly well aware of, as Switch features a CPU boost mode specifically for improving loading times. It's a clear win for a fixed-function block like this, as modern game engines are constantly loading (and decompressing) assets from disk, and doing so on dedicated hardware will be more efficient in terms of transistors and power than throwing more CPU cores at it. Sony have gone this route with the PS5, where they've licensed both an algorithm and the hardware IP from an external source, whereas Nvidia seem to have done this in-house.

I think this is a pretty good sign for the next Switch. I've been saying for quite a while that dedicated decompression hardware would be a sensible thing to add to the new console, so it's good to see Nintendo and Nvidia thinking along the same lines. Of course it doesn't mean anything like PS5 load speeds, and we're still going to be limited by the read speeds of game cards, internal storage and removable storage, but it suggests Nintendo are serious about improving load speeds and decreasing CPU overhead of asset streaming. Which bodes well for them also using faster storage media too.
The external source you’re referring to with PS5 is Oodle, yeah? The data decompression on that was impressive, but the real big win there was the texture compression tech, hope Nvidia offers something like that.
 
So, it looks like Drake has a File Decompression Engine, a new piece of hardware which isn't on Orin.

A while ago after it was found out that Drake is using an 8 core CPU from a Linux commit message, I had a browse around myself and found some differences in some of the hardware blocks between Orin and Drake. Most of these are pretty understandable (removing automotive-focussed hardware), but one of them has stuck in my brain for a while, which is that there's a new block on Drake that's not on Orin, labelled FDE. I find this interesting because if it's on Drake but not on Orin, it would likely be something requested by Nintendo, or at least something that's of particular usefulness for a games console.

I just found a commit which solves the conundrum. The commit message simply reads:



I've also found a LinkedIn profile of a Nvidia employee which mentions FDE being "File Decompression Engine for games". I won't share it here, as I don't really feel comfortable posting links to random people's social media on public forums, but you can find it easily enough if you search. In any case, if the block is on Drake but not Orin, then it's not a huge leap to say it's for games.

A File Decompression Engine makes a lot of sense for a games console. Decompressing files as they're read from disk has a significant CPU overhead, something Nintendo are clearly well aware of, as Switch features a CPU boost mode specifically for improving loading times. It's a clear win for a fixed-function block like this, as modern game engines are constantly loading (and decompressing) assets from disk, and doing so on dedicated hardware will be more efficient in terms of transistors and power than throwing more CPU cores at it. Sony have gone this route with the PS5, where they've licensed both an algorithm and the hardware IP from an external source, whereas Nvidia seem to have done this in-house.

I think this is a pretty good sign for the next Switch. I've been saying for quite a while that dedicated decompression hardware would be a sensible thing to add to the new console, so it's good to see Nintendo and Nvidia thinking along the same lines. Of course it doesn't mean anything like PS5 load speeds, and we're still going to be limited by the read speeds of game cards, internal storage and removable storage, but it suggests Nintendo are serious about improving load speeds and decreasing CPU overhead of asset streaming. Which bodes well for them also using faster storage media too.
Makes all the sense in the world given that every other piece og gaming hardware is prioritizing dedicated decompression. Nintendo might also leverage Nvidia's open source format gdeflate
 
Kingdom Battle is arguably the best looking Switch game. I dunno why people leave it off their list. Sparks of Hope going for full camera control in a bigger RPG styled world required some sacrafices. That said it still looks like there is a bunch of environmental detail and geometry present. Side grade seems like the correct term. Would have been nice to see some reconstruction techniques though similar to XC3 or so TAA.

The game still looks great. Just couldn't pack in the same detail with the expanded scope. Bummer but I mean again, Kingdom Battle has imo always been the best looking Switch game. The amount of geometry in that game was crazy and it along with Luigi's Mansion 3 are the Switch games that look a lot closer to straight PS4/Xbox One games with little compromises.
Monster Hunter Rise is in with a shout for best looking Switch game.
 
Is it true that the T239 is a T234 but with a downgrade in its performance or power?
It's a lot more complicated than that. The two chips are related, but they're configured very differently and tuned for different goals. T234 is likely more capable overall, but it's not configured super well for a game console, especially one that has to run off a battery. T239 sheds a lot of the more automotive focused extras and, as far as we can tell, is configured closer to desktop Ampere.
So, it looks like Drake has a File Decompression Engine, a new piece of hardware which isn't on Orin.

A while ago after it was found out that Drake is using an 8 core CPU from a Linux commit message, I had a browse around myself and found some differences in some of the hardware blocks between Orin and Drake. Most of these are pretty understandable (removing automotive-focussed hardware), but one of them has stuck in my brain for a while, which is that there's a new block on Drake that's not on Orin, labelled FDE. I find this interesting because if it's on Drake but not on Orin, it would likely be something requested by Nintendo, or at least something that's of particular usefulness for a games console.

I just found a commit which solves the conundrum. The commit message simply reads:



I've also found a LinkedIn profile of a Nvidia employee which mentions FDE being "File Decompression Engine for games". I won't share it here, as I don't really feel comfortable posting links to random people's social media on public forums, but you can find it easily enough if you search. In any case, if the block is on Drake but not Orin, then it's not a huge leap to say it's for games.

A File Decompression Engine makes a lot of sense for a games console. Decompressing files as they're read from disk has a significant CPU overhead, something Nintendo are clearly well aware of, as Switch features a CPU boost mode specifically for improving loading times. It's a clear win for a fixed-function block like this, as modern game engines are constantly loading (and decompressing) assets from disk, and doing so on dedicated hardware will be more efficient in terms of transistors and power than throwing more CPU cores at it. Sony have gone this route with the PS5, where they've licensed both an algorithm and the hardware IP from an external source, whereas Nvidia seem to have done this in-house.

I think this is a pretty good sign for the next Switch. I've been saying for quite a while that dedicated decompression hardware would be a sensible thing to add to the new console, so it's good to see Nintendo and Nvidia thinking along the same lines. Of course it doesn't mean anything like PS5 load speeds, and we're still going to be limited by the read speeds of game cards, internal storage and removable storage, but it suggests Nintendo are serious about improving load speeds and decreasing CPU overhead of asset streaming. Which bodes well for them also using faster storage media too.
Do we know if this is the same decompression that falls under the RTX IO brand on desktop?

Regardless, one does get the impression that a lot of loading stuff on the current Switch is probably CPU bound, so this is definitely a smart inclusion.
 
So, it looks like Drake has a File Decompression Engine, a new piece of hardware which isn't on Orin.
Nice sleuthing! Transparent decompression was my guess for the FDE, good to know I’m not totally off base

Do we know if this is the same decompression that falls under the RTX IO brand on desktop?
It is not. RTX IO is really three technologies working together, the decompression portion is “merely” CUDA accelerated, not a specialized fixed function block. In the case of RTX IO, there is a specific win if the GPU performs the decompression, even if the GPU isn’t faster than the CPU.
 
0
DF tech review of Mario+Rabbids : Sparks of hope:


I can't help but feel like I am looking at a Switch port of a PS4 game(I am not talking about in comparison with the first game, even though Sparks of hope with its extended areas could have also looked like that on better hardware). But, like the game was intended to run on better hardware(but couldn't). Remember Drake could have been launched by now per Nate and Bloomberg and TOTK was first dated as a 2022 release.

I can't be the only one feeling that way...

people have been saying that for years
 
RTX IO discussion, because it comes up a bit, and because I myself was confused a bit till I dug into it.

Say you've got a bunch of textures on disk, and you need to load them up to render them.

The files are on the file system, scattered about. You fire up a read request for each one, which reads from (in this example) an SSD pretty fast. You then block the thread while you wait for the data to come off disk.

Then you decompress the image using the CPU, from whatever format it currently is, into raw bitmapped data which your GPU can use.

Then you copy that data from main memory into VRAM. The security model gets in your way a bit here, because unprivileged code (the game) isn't allowed direct access to VRAM, so you have to go through the driver.

Once done, you can now perform whatever GPU operations on those textures you want.

Say you've got a bunch of textures on disk, and you need to load them up to render them.

The files are on the file system, scattered about. DirectStorage provides a special API for reading files that lets you set up "queues" of file data. On the CPU side you add a bunch of files to the queue, and the GPU reads out of the queue. In the middle, the storage driver, instead of executing a bunch of random reads on the SSD, they can look at all the files you want to eventually get, and batches them into a single, sequential read request, maximizing disk read speed.

DirectStorage works with the kernel to bypass the usual security model, so that, effectively, the data never passes through main memory at all but instead goes straight from disk to the GPU.

However, that's still not enough. Ordinarily the GPU can only work with decompressed data. RTX IO fixes that problem by providing decompression that works transparently on the CUDA cores. Even when that is slower than doing it on the CPU, it's a net win because you're just copying a small amount of data to the GPU rather than a large amount of decompressed data, and the GPU decompression can happen when the GPU is usually idle anyway

In the case of a console like the Switch the OS security model and the physical hardware are totally different. Games run with driver level privileges for the GPU, so there is no need for a kernel workaround to get to the GPU, and there aren't separate memory pools for the GPU vs the CPU, so the copy is already avoided.

The RTX IO solution only works for data the GPU will eventually work on. Potentially a fixed-function decompression block on a console not only could outperform RTX IO's CUDA only decompression algorithms, but in theory it can support formats that the CPU would work on - for example level data, sound formats, even text and compressed game script.
 
0
Is it true that the T239 is a T234 but with a downgrade in its performance or power?
Yes. It’s a downgrade as a machine learning, mini supercomputer meant for cars.

No because it’s an upgrade over Orin as an actual gaming device.

Which Orin isn’t meant for or presumably good at.
 
I forgot to comment myself on this: the SoC is perfect and has checked every personal checklist, I can die peacefully now. I have zero real qualms now.

CPU? Perfect upgrade. GPU? perfect upgrade. Decompression block? Perfect upgrade. Memory? Perfect upgrade. OLED display? Perfect upgrade. HDMI 2 or later? Perfect upgrade.

I have no qualms and it’s a perfect upgrade in every facet over the switch. This will last me until 2030 or whatever, I don’t care.

The SoC is a perfect upgrade…..


Actually there’s one thing I want to know, will this have an SLC? That’s still a mystery to me, this wouldn’t really be that apparent or visible, but it would be a Bandwidth requirement reduction, and efficiency increase.

4-6MB SLC would be pretty sweet. Orin has a 4MB SLC, so maybe this remained?
 
Last edited:
Is it possible for drake to top out at about 1A current like the 2019/OLED switch on full load without severely underclocking the CPU/GPU?
If it tops out at about 1.5A but the battery is upgraded from a ~4100mAh to a ~6000mAh then we'd have the same ~4 hours minimum on full blast like the 2019/OLED model.
The question being, how exactly feasible is the nearly 2000mAh upgrade.
I mean, the switch's battery compartment seems to be a bit small but I believe at the very least, 5000~5500 should be possible (judging from some cheap android phones).
They could maybe even do it like smartphone companies and make the battery taller, but thinner and re-arrange the other internals.
 
Is it possible for drake to top out at about 1A current like the 2019/OLED switch on full load without severely underclocking the CPU/GPU?
If it tops out at about 1.5A but the battery is upgraded from a ~4100mAh to a ~6000mAh then we'd have the same ~4 hours minimum on full blast like the 2019/OLED model.
The question being, how exactly feasible is the nearly 2000mAh upgrade.
I mean, the switch's battery compartment seems to be a bit small but I believe at the very least, 5000~5500 should be possible (judging from some cheap android phones).
They could maybe even do it like smartphone companies and make the battery taller, but thinner and re-arrange the other internals.
I can see Nintendo going with a 5k mAh battery if the situation demands for it. Plus, it should be a bit smaller than the current switch battery at that capacity. Saving space would be nice.


Would mean that they can also stretch a bit more for TDP if need be. And in that case you’d get minimum of ~3.5-3.75Hrs being on a better node process, not subject to leak issues of the 20nm. So, it would be a steps back from the OLED and v2, but an upgrade over the V1 and the lite perhaps.
 
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