Nintendo Switch Dev Kit Stats Leaked? Cortex A57, 4GB RAM, 32GB Storage, Multi-Touch.

Status
Not open for further replies.
I didn't click the link, but thanks for the heads up.

Still, that guy is generally right, I think. I think he's the WSJ Japan tech guy. With as little as we know, I don't doubt they know something, but it kinda is at odds with other rumors so far. His recent Tweet seems to imply that they'll adjust the clocks at a later time. Is what they're doing now a temporary solution while they test out real world use and based on that they'll do a firmware update and adjust base clock speeds?
That is a very reasonable explanation for why these clocks are so low in the first place. As all the tech people here have explained the switch should easily be able to handle higher clocks.
 
That is a very reasonable explanation for why these clocks are so low in the first place. As all the tech people here have explained the switch should easily be able to handle higher clocks.

There is no extra cooling in the dock as far as we know (from the patents) and based on devices like the pixel c, we can assume that Switch in portable mode could upclock to the same clocks as when docked. Though they would be better off clocking up the CPU.

I'd suggest a 25% overclock of both the CPU and GPU in switch if they would do it, it is a safe overclocking and would give 205gflops in portable, 512gflops when docked and 4 1.25ghz A57 cores should be close to 5 PS4 cores, so developers would have less optimization to do when porting games.

We should not expect an upclock, but if Switch needs more power, Nintendo can provide it with the above solution.
 
There is no extra cooling in the dock as far as we know (from the patents) and based on devices like the pixel c, we can assume that Switch in portable mode could upclock to the same clocks as when docked. Though they would be better off clocking up the CPU.

I'd suggest a 25% overclock of both the CPU and GPU in switch if they would do it, it is a safe overclocking and would give 205gflops in portable, 512gflops when docked and 4 1.25ghz A57 cores should be close to 5 PS4 cores, so developers would have less optimization to do when porting games.

We should not expect an upclock, but if Switch needs more power, Nintendo can provide it with the above solution.

At the moment, the current situation doesn't need cooling, right? So that would give them some room to change the clock speeds over time depending on what the average use for the device is in real world terms. We still don't know in what way the SoC was customized, either. I'm just trying to make sense of his last Tweet and changing the clocks via firmware at some point makes the most sense to me.
 
I agree. I wonder why he would put his reputation on the line like this? Unless he knows this guy and trusts his information. Either way, I guess we'll know for sure in a couple of weeks.

Other than Kingsnake reminding us Yusuda (Analyst) is the same guy that speculated that Switch would have 16nmFF.

I'm still wondering why he'd speculate about Displayport?

https://en.m.wikipedia.org/wiki/DisplayPort

Ace Research Institute analyst Yasuda: NintendoSwitch uses Displayport over USB-C; internal bus speed at 5Gbps, faster than 3DS's 128Mbps

https://twitter.com/mochi_wsj/status/811850362785148928

Would it be for the same reason I said? To avoid paying licensing and royalties to HDMI? Can they get around that using an HDMI adapter?

Also, wouldn't bus speed be referring to Hz not bits per second as its unit? I guess it can still be inferred the same way as it's per second.

The only 128 value I recognise is the FCRAM for the 3DS which is just the amount of RAM 128MB.

Edit: Maybe he's saying Switch has 5 GB of RAM? /jk
 
Speaks for the success of the Nvidia Shield that no one here tries to clock down it to the speculated specs and runs UE demos through Vulkan.

I was thinking about this the other day... clearly not many people have a Shield TV, although it's not clear how easy it would be to downclock if they did. I think the Android overhead would make any comparison pointless though.


Weren't those based on an analyst reading a patent? Was that ever clarified? It could be new patents published in Japan.

There is no extra cooling in the dock as far as we know (from the patents) and based on devices like the pixel c, we can assume that Switch in portable mode could upclock to the same clocks as when docked. Though they would be better off clocking up the CPU.

To be fair Laura Dale did report that there is an extra fan in the dock, and this was much more recently than the patent was filed. It's very possible that they added that into the final hardware since June.
 
Other than Kingsnake reminding us Yusuda (Analyst) is the same guy that speculated that Switch would have 16nmFF.

I'm still wondering why he'd speculate about Displayport?

https://en.m.wikipedia.org/wiki/DisplayPort



https://twitter.com/mochi_wsj/status/811850362785148928

Would it be for the same reason I said? To avoid paying licensing and royalties to HDMI? Can they get around that using an HDMI adapter?

Also, wouldn't bus speed be referring to Hz not bits per second as its unit? I guess it can still be inferred the same way as it's per second.

The only 128 value I recognise is the FCRAM for the 3DS which is just the amount of RAM 128MB.

Edit: Maybe he's saying Switch has 5 GB of RAM? /jk

But we saw the USB-C port, didn't we? Plus the cable accessories EBGames Aus put up a while ago.
 
But we saw the USB-C port, didn't we? Plus the cable accessories EBGames Aus put up a while ago.

What I think he means for Displayport over USB-C means that the Switch still uses a USB-C port, it's that the dock has a Displayport where we would've expected an HDMI port.

This is from either his speculation or from reading a patent that we haven't seen because it's only been published in Japan and hasn't appeared in the US yet.

Just speculating this.

Edit: Referring to this tweet about patents.

Ace Yasuda: patent filings show Nintendo Switch's architecture good for open-world games. End!

https://twitter.com/mochi_wsj/status/811851248106283008
 
What I think he means for Displayport over USB-C means that the Switch still uses a USB-C port, it's that the dock has a Displayport where we would've expected an HDMI port.

This is from either his speculation or from reading a patent that we haven't seen because it's only been published in Japan and hasn't appeared in the US yet.

Just speculating this.

Edit: Referring to this tweet about patents.

Is there confirmation that there are newly published Japanese patents? There are ways to view/translate those if we can find out some information (application number, inventor name, etc.)
 
Huh, I was under the impression HDMI was still completely sufficient for consoles? Especially one that doesn't target 4k or anything like that.

And since it has to go DisplayPort -> HDMI on your TV anyway, wouldn't you lose the benefits that way?

Just seems really random for DisplayPort to be thrown into the ring.
 
Is there confirmation that there are newly published Japanese patents? There are ways to view/translate those if we can find out some information (application number, inventor name, etc.)

That's why I said it's just speculation.

Now that I realise it:

Ace Research Institute analyst Yasuda: NintendoSwitch uses Displayport over USB-C; internal bus speed at 5Gbps, faster than 3DS's 128Mbps
Ace Yasuda: patent filings show Nintendo Switch's architecture good for open-world games. End!

This was a continuing sentence, but the technical jargon sounded wrong.

I had it wrong, it's not a continuing sentence but where the hell would he get the idea about the architecture being good for open world games from a patent?

Bus speed should be referring to Hz, the only reference I know of 3DS having 128 of something is 128MB of FCRAM.

So he is most likely saying the Switch has 5GB of RAM which makes it "good for open world games".

Problem is, Nintendo tech specs like the above don't show up in patents do they?

Otherwise it sounds more like it's his speculation just like the 16nmFF speculation from Yusuda.
 
"Made for open world games" is all you need to read to see the analyst is just that, an analyst, doesn't know shit and predict things according to what makes most sense atm.

Open world games are all the rage these days? surely this thing is made for them!
 
Even with the current figures I am still more worried with Nintendo choosing a crappy panel than with polycount or shaders. My main graphical problem with the Wii U off-screen was not standard definition or PS3-like graphics but poor colour and visibility.
 
This isn't true, low spec gamer uses utterly ancient stuff and look at the results.

Notice what vulkan does?

If you don't like dog turds fine, but millions of consumers aren't really as picky as certain console gamers would like to argue. When something like minecraft or event GTA5 can dominate vs the current exceptional climate of heavy hitting graphical titles this argument needs to stop.

The priority will be what will nintendo realistically do to get 3rd parties involved. No point in putting this in a PM, but if Nvidia/AMd works with developers on games on PC I see no reason why nintendo as a platform holder can't grow up in this area when those two companies along with sony and ms do a pretty fucking bang up job. They should be getting ports up and running and if they don't have people who do better than some dude doing youtube videos or tweakers like myself they should be fired promptly. Shogmaster is right to ask for nintendo balls up and just say this is the ultimate portable.

You don't need magic sauce you just need to use vulkan and do a solid job on optimizing. Doom be it low spec machines or not using vulkan has some serious gains if we are talking performance.

Your video just proves my point it looks like dog turds.
 
Yeah, patents don't mention actual specs at all.

I had it wrong, it's not a continuing sentence but where the hell would he get the idea about the architecture being good for open world games from a patent?

Better to take all this as speculation rather than fact. Maybe we'll see a new WSJ article pop up? lol
 
I had it wrong, it's not a continuing sentence but where the hell would he get the idea about the architecture being good for open world games from a patent?
The sentence makes no sense either way. Architecture means nothing for open world games.
 
unless the switch has some sort of secret sauce. At 400 TF, only 4gb ram, low clocked cpu, very low memory bandwidth it is not getting any modern games from this console generation unless they look like dog turds. There will probably be a lot of last gen ports

Like Rise of the Tomb Raider looking like a dog turd on the Xbox 360. Wait, it didn't, it arguably looked better than Tomb Raider 2013. There's only one set of numbers in the way of the Switch getting ports and it's sales.
 
The sentence makes no sense either way. Architecture means nothing for open world games.

Yep, that's why I didn't get what he was trying to say. I'll just chalk it up to him being Japanese and English is not his first language. At least the "open world" part. Open worlds have been part of gaming for a LONG time so what does it mean to be "made for open world"? Nothing to worry about, but like someone said, I wouldn't be surprised if there was a WSJ article sometime soon.
 
The sentence makes no sense either way. Architecture means nothing for open world games.

And as mentioned, bus speed sounded wrong.

It's like he used all this technical jargon in the wrong way.

"The Xbox One has fifty gigabytes of CPU and runs a hundred kilometres per second on RAM. Its architecture makes it good for walking simulators and yoga instructors."
 
That's why I said it's just speculation.

Now that I realise it:

This was a continuing sentence, but the technical jargon sounded wrong.

I had it wrong, it's not a continuing sentence but where the hell would he get the idea about the architecture being good for open world games from a patent?

Bus speed should be referring to Hz, the only reference I know of 3DS having 128 of something is 128MB of FCRAM.

So he is most likely saying the Switch has 5GB of RAM which makes it "good for open world games".

Problem is, Nintendo tech specs like the above don't show up in patents do they?

Otherwise it sounds more like it's his speculation just like the 16nmFF speculation from Yusuda.

Yeah, patents don't mention actual specs at all.

Patents can most certainly mention specs, but they would almost definitely be giving a range in specs, and likely just as an example. Saying something like:

Typical Patent Language Example said:
The interactive game device 10 also includes a memory module 15 which can, for example, contain between 2 and 8 gigabytes of LPDDR4 memory.

If the patent has this type of spec discussion then the analyst likely just looked at the top of the range and assumed that would be what they are going for, which is a very faulty assumption. And examples in particular mean absolutely nothing for the patent coverage or what the final device will have. The discussion about the bus speed seems like he got something confused though, yeah.

How can a patent tell if it's good for open world games?

It's very possible that the patent explicitly mentions, as an example, that it is useful or capable of running demanding open world games or something like that. Again, basing this type of analysis on a patent is pretty poor analysis, but it's certainly possible things like this could be in some newly published patents.
 
And as mentioned, bus speed sounded wrong.

It's like he used all this technical jargon in the wrong way.

"The Xbox One has fifty gigabytes of CPU and runs a hundred kilometres per second on RAM. Its architecture makes it good for walking simulators and yoga instructors."
Maybe he meant bus bandwidth, it was the right unit for that. But 5GB/s sounds incredibly low based on earlier discussion in this thread. No idea where he got that number.
 
Maybe he meant bus bandwidth, it was the right unit for that. But 5GB/s sounds incredibly low based on earlier discussion in this thread. No idea where he got that number.

No you see the switch CPU will obviously be capable of up to 200MPH on land and 300MPH in the air while docked. Making it the fastest CPU available :P
 
Other than Kingsnake reminding us Yusuda (Analyst) is the same guy that speculated that Switch would have 16nmFF.

I'm still wondering why he'd speculate about Displayport?

https://en.m.wikipedia.org/wiki/DisplayPort



https://twitter.com/mochi_wsj/status/811850362785148928

Would it be for the same reason I said? To avoid paying licensing and royalties to HDMI? Can they get around that using an HDMI adapter?

Also, wouldn't bus speed be referring to Hz not bits per second as its unit? I guess it can still be inferred the same way as it's per second.

The only 128 value I recognise is the FCRAM for the 3DS which is just the amount of RAM 128MB.

Edit: Maybe he's saying Switch has 5 GB of RAM? /jk
The one context where those bus speeds make sense for 'open world games' is if he's talking of the bus to the game medium - medium BW (and latency) is paramount to the success of streaming asset tech.

For reference, 5Gbps is not that far from SATA3's 6Gbps.
 
The one context where those bus speeds make sense for 'open world games' is if he's talking of the bus to the game medium - the medium BW (and latency) is paramount to the success of streaming asset tech.

Would it even be possible to get game card speeds anywhere close to 5GB/s? Isn't that pretty absurdly fast for that type of game medium?

If that's what he means then people who go digital will have serious disadvantages in some games...
 
Would it even be possible to get game card speeds anywhere close to 5GB/s? Isn't that pretty absurdly fast for that type of game medium?

If that's what he means then people who go digital will have serious disadvantages in some games...
Serial buses are measured in gigabits ; )
 
Other than Kingsnake reminding us Yusuda (Analyst) is the same guy that speculated that Switch would have 16nmFF.

I'm still wondering why he'd speculate about Displayport?

https://en.m.wikipedia.org/wiki/DisplayPort



https://twitter.com/mochi_wsj/status/811850362785148928

Would it be for the same reason I said? To avoid paying licensing and royalties to HDMI? Can they get around that using an HDMI adapter?

Also, wouldn't bus speed be referring to Hz not bits per second as its unit? I guess it can still be inferred the same way as it's per second.

The only 128 value I recognise is the FCRAM for the 3DS which is just the amount of RAM 128MB.

Edit: Maybe he's saying Switch has 5 GB of RAM? /jk

From http://www.displayport.org/what-is-displayport-over-usb-c/

Advantages of DisplayPort over USB-C

  • Full DisplayPort audio/video (A/V) performance (up to 8K at 60Hz)
  • SuperSpeed USB (USB 3.1) data
  • Up to 100 watts of power over a single cable
  • Reversible plug orientation and cable direction
  • Backwards compatibility to VGA, DVI with adaptors
  • Supports HDMI 2.0a and is backwards compatible with previous versions

I'd say the big advantage of DisplayPort over USB-C vs HDMI over USB-C would be the points in bold. HDMI over USB-C is limited to HDMI 1.4b spec (4k30fps, no HDR). I'd imagine they'd still have to pay the same HDMI royalty, one way or the other, but it'd allow them to sidestep the arbitrary HDMI limitation and achieve Netflix parity with the other systems.
 
Nah, that was to do with their insistence that the media is lying and that Nintendo will come out with something greater than what is reported from AMD and DMP while making up shit to prove it.

If you want to talk about a console more powerful than the competition or a 4 TFLOPS peripheral attachment, go ahead.

However, if you're making the claim that you expect Nintendo to actually make a powerful console when Yamauchi is no longer around and the GameCube proved that the audience isn't there for those that want a system more powerful than the competition. Then it gives Nintendo less reason to go with a Scorpio killer unless they need power for a particular reason like VR, we've already seen proof now with the Switch.

Switch is currently estimated to not even be half as powerful as an Xbox One, that shows Nintendo are fine at the level that they want to handle games at, they don't care about brute forcing everything with raw power to be on the same level as their competitors.

Fair enough, I understand keeping expectations in check and you turned out to be absolutely correct, I just don't see the harm in discussing the possibility of Nintendo offering customers the option of more power if they so desire it.

I don't think they will go with SCD's personally but I do think that Switch is just one in a long line of devices all built around the same architecture / OS. IMO next up will be a standalone console and if they use the Switch architecture as a base they can remove the joycons, the screen etc and boost the chips up to 2GHz on the CPU and around 1.5GHz on the GPU, throw in another 4GB's of RAM, a larger HDD, put it all in a small Gamecube like box and they could sell it for $199 in late 2019.

I see a very cheap dedicated handheld coming after that in 2020 (a slimmed down Switch at 16nm) and then maybe even a tablet aimed at very young children (depending on how successful their own mobile games are).

Three or four cheaply priced devices which appeals to different demographics is the future of Nintendo hardware success imo. Time will tell but I wouldn't rule out them mentioning more devices in Jan so that they are not seen to blindside customers when more are eventually announced.
 
So we'd need to dig up info of the bus bandwidth of the 3DS cartridge to see if he correctly referenced it using the 128Mbps value.

https://www.3dbrew.org/wiki/Gamecards
16.6MHz * 8 data bits = 132.8Mbps, I'd say close enough for a twitter post.

Ah foiled again by acronyms.

SD cards top out at ~300MB/s which is what, like 2.4Gb/s? 5Gb/s would still be rather high, right?
Correct. As already mentioned, SATA3 is 6Gbps, so if that BW is what I think it is, Switch's gamecarts will be able to achieve (read: not necessarily for all cart sizes) speeds just shy of SATA3.
 
The one context where those bus speeds make sense for 'open world games' is if he's talking of the bus to the game medium - medium BW (and latency) is paramount to the success of streaming asset tech.

For reference, 5Gbps is not that far from SATA3's 6Gbps.

He's probably speculating on the USB-C bus speed being 5Gbps, or USB 3.0 rather than USB 3.1 (10Gbps) or thunderbolt 3 (40Gbps) It would mean that data transfer from the dock to the tablet would be limited to 5Gbps, which would limit an eGPU. He is speculating on the safest bet here. I think the limit on bus if true would mean that you'd get express port 2 performance, so at best 2TFLOPs (?) for a theoretical expansion dock.
 
Really wish they went with A72's.

How much of a difference would that make given there are still 4 cores at around 1 GHz? Genuinely asking, I'm not too knowledgeable on CPU cores.

Correct. As already mentioned, SATA3 is 6Gbps, so if that BW is what I think it is, Switch's gamecarts will be able to achieve (read: not necessary on all carts) speeds just shy of SATA3.

Gotcha, thanks.

If this does pan out I am curious what it will mean for digital purchases. Will developers actually be able to take advantage of game card read speed as an actual factor in running the game?
 
From http://www.displayport.org/what-is-displayport-over-usb-c/



I'd say the big advantage of DisplayPort over USB-C vs HDMI over USB-C would be the points in bold. HDMI over USB-C is limited to HDMI 1.4b spec (4k30fps, no HDR). I'd imagine they'd still have to pay the same HDMI royalty, one way or the other, but it'd allow them to sidestep the arbitrary HDMI limitation and achieve Netflix parity with the other systems.

Yeah, that makes sense since I wondered why HDMI Alt mode wasn't up to the lastest model/spec yet. Hence using DisplayPort as an alternative.

https://www.3dbrew.org/wiki/Gamecards
16.6MHz * 8 data bits = 132.8Mbps, I'd say close enough for a twitter post.


Correct. As already mentioned, SATA3 is 6Gbps, so if that BW is what I think it is, Switch's gamecarts will be able to achieve (read: not necessary for all cart sizes) speeds just shy of SATA3.

So then 5Gbps for a Switch game card makes sense with his claims of Switch being good for open world games since it would load things fast enough.
 
From http://www.displayport.org/what-is-displayport-over-usb-c/



I'd say the big advantage of DisplayPort over USB-C vs HDMI over USB-C would be the points in bold. HDMI over USB-C is limited to HDMI 1.4b spec (4k30fps, no HDR). I'd imagine they'd still have to pay the same HDMI royalty, one way or the other, but it'd allow them to sidestep the arbitrary HDMI limitation and achieve Netflix parity with the other systems.

The ports and technology mixing has become such a mess with USB C. So it could possibly use the physical USB C port, use the DisplayPort standard and USB 3.1 for data/video/power protocol handling and then it gets split in the dock into USB A ports and a HDMI port. So that way you can deliver 15W+ of power, transfer data and a high bit rate video stream while still supporting a HDMI compatibility with tv's. No idea if they would actually go this way, but it looks like a possibility, even though it might be slightly more expensive to implement the hardware in the dock. The end user wouldn't notice anything of it though, they'll just plug in their ports and be done with it.
 
From http://www.displayport.org/what-is-displayport-over-usb-c/



I'd say the big advantage of DisplayPort over USB-C vs HDMI over USB-C would be the points in bold. HDMI over USB-C is limited to HDMI 1.4b spec (4k30fps, no HDR). I'd imigine they'd still have to pay the same HDMI royalty regardless.

No, more like " Why would you think Nintendo would have DP on whats supposed to be a home console when DP is a technology mostly relegated to computer monitors." Up till 2015 DP was not available in tvs. I still dont think tvs have that input all that often, if any at all.

Like, I cant even believe people are suggesting that its a royalty thing. HDMI royalties are like 5 cents per device after the initial yearly subscription. A pittance. It would cost more to add a dp -> hdmi converter than just using a straight up HDMI port. And there is no technical reason why you'd need the extra horsepower DP has over HDMI for a console, much less the switch of all things.
 
How much of a difference would that make given there are still 4 cores at around 1 GHz? Genuinely asking, I'm not too knowledgeable on CPU cores.



Gotcha, thanks.

If this does pan out I am curious what it will mean for digital purchases. Will developers actually be able to take advantage of game card read speed as an actual factor in running the game?

1. A72 cores could be clocked higher for the same power draw.
2. They get 11% to 20% better performance per clock iirc.

So even at 1Ghz 4 A72 cores would achieve nearly the same performance as 5 PS4 cpu cores, but it is very likely we would have seen a clock bump if they went with A72 (assuming that the data sheet is correct and they are going with A57) so 1.2ghz with that clock could have been fairly likely, putting the CPU right around PS4's 6 cores (not counting the half core)
 
1. A72 cores could be clocked higher for the same power draw.
2. They get 11% to 20% better performance per clock iirc.

So even at 1Ghz 4 A72 cores would achieve nearly the same performance as 5 PS4 cpu cores, but it is very likely we would have seen a clock bump if they went with A72 (assuming that the data sheet is correct and they are going with A57) so 1.2ghz with that clock could have been fairly likely, putting the CPU right around PS4's 6 cores (not counting the half core)

Wait...has it been 100% confirmed that they're using A57s? (I've been out for a couple of days so I might be a bit behind in what's been confirmed.)
 
Wait...has it been 100% confirmed that they're using A57s? (I've been out for a couple of days so I might be a bit behind in what's been confirmed.)

No, we are speculating on data sheet, but I'll go one part further with A72 and 16nm, had this been the case with Switch instead of A57 on 20nm (which we assume is the case) They could have consumed the same power and clocked 4 A72 cores to 1.7ghz.
 
No, more like " Why would you think Nintendo would have DP on whats supposed to be a home console when DP is a technology mostly relegated to computer monitors." Up till 2015 DP was not available in tvs. I still dont think tvs have that input all that often, if any at all.

Like, I cant even believe people are suggesting that its a royalty thing. HDMI royalties are like 5 cents per device after the initial yearly subscription. A pittance. It would cost more to add a dp -> hdmi converter than just using a straight up HDMI port. And there is no technical reason why you'd need the extra horsepower DP has over HDMI for a console, much less the switch of all things.

I'm not saying they'll do it, just that if they do, that would be the most obvious reason why. Like I said, it'd be more for Netflix and video streaming than gaming (though HDR could still be beneficial for gaming at any resolution).

Also, the Docks going to require active circuitry in it anyway, to split the IO. Would it not be possible to use the same chip to do the DVI to HDMI conversion?

But yeah, it's Nintendo, so I don't really expect them to do any of that.
 
Fair enough, I understand keeping expectations in check and you turned out to be absolutely correct, I just don't see the harm in discussing the possibility of Nintendo offering customers the option of more power if they so desire it.

I don't think they will go with SCD's personally but I do think that Switch is just one in a long line of devices all built around the same architecture / OS. IMO next up will be a standalone console and if they use the Switch architecture as a base they can remove the joycons, the screen etc and boost the chips up to 2GHz on the CPU and around 1.5GHz on the GPU, throw in another 4GB's of RAM, a larger HDD, put it all in a small Gamecube like box and they could sell it for $199 in late 2019.

I see a very cheap dedicated handheld coming after that in 2020 (a slimmed down Switch at 16nm) and then maybe even a tablet aimed at very young children (depending on how successful their own mobile games are).

Three or four cheaply priced devices which appeals to different demographics is the future of Nintendo hardware success imo. Time will tell but I wouldn't rule out them mentioning more devices in Jan so that they are not seen to blindside customers when more are eventually announced.

That's nothing controversial, I already speculated long before that if they were going to use multiple form factors, then there could be a PS Vita TV equivalent and a handheld only equivalent.

Considering how things are underclocked, it is more likely easier to see the console only version becoming upclocked. Whether it is upclocked more than the Switch, I don't know.

It just gets into the wish list area if someone said something like, "I expect Nintendo will bring a home console in 2017/2018 and it will be stronger than a PS4."

That definitely happened with the Eurogamer article about the CPU/GPU clocks.

They took probably around 3+ years producing the Switch. It is easier to predict a revision of the Switch in a new form factor than to suddenly run Project Scorpio in less than 2 years.
 
I'm no Jeff Rigby(!) But I figured the dock would be dealing with a distinct HDMI connection with the TV and the signal from the tablet to the dock sent by whatever works best, if display port over usb-c is more mature it makes sense to me.
This is probably from my irrational fear of handshaking issues tbh.
 
So I just had a thought. Since, if a game publisher decides that "the PC port of Game X is going to have Nvidia specific features", then Nvidia themselves will send an engineer on-site to make sure that the developers can squeeze every last drop of performance from Nvidia GPUs, do you think that Nvidia would do the same thing if publishers announced that "game Y is going to have a Switch port"? Please note that the whole "Nvidia sending engineers out to assist in porting for Nvidia specific features" thing is a recent development.
I hope so.
 
Status
Not open for further replies.
Top Bottom