Switch 2 Speculation

...

So Sony and MS who both have new home consoles wouldn't need to design their portable to work in a dock. They could simply make the distinction that this other product is for gaming on the go

I think Nintendo should go mobile only for the futur. They're not a player in the home console business anymore (sadly). They already released a mobile only switch version...
 
The new unannounced Switch model is still using the Mariko SoC. This should put to rest the idea that GPU side emulation is somehow even a viable path for software compatibility. I don't think Nvidia plans on using the best possible transistor technology for low-end projects either since Orin will be produced on Samsung's 8nm process which will be available next year ? I think Nintendo might be stranded for their next successor since Nvidia places the least amount of effort for lower power SoCs ...

I think the Samsung partnership rumor could actually be gaining some real ground since they can effectively subsidize the development of low power integrated circuits when they have a sustainable customer base buying their phones ...
 
The new unannounced Switch model is still using the Mariko SoC. This should put to rest the idea that GPU side emulation is somehow even a viable path for software compatibility. I don't think Nvidia plans on using the best possible transistor technology for low-end projects either since Orin will be produced on Samsung's 8nm process which will be available next year ? I think Nintendo might be stranded for their next successor since Nvidia places the least amount of effort for lower power SoCs ....
nvidia always try to put old overpriced hardware to console (rsx in ps3, old tegra in switch) probably thats one of the reason it's no longer partner for microsoft or sony
 
Why enter an unwinnable territory when they already have xcloud available for every android mobile without any additional cost?
How does xcloud replace a handheld gaming device?

I want a handheld to be able to play games when traveling by plane or train or when I go on a short vacation to a rural area with poor cellphone reception or on a long vacation to a caribbean or middle-eastern or southeast asian country.
Xcloud, Stadia, PS Now etc. would all be absolutely worthless for that.

If I only wanted a handheld to play games on my house's toilet or my bed I'd already be served with Steam In-Home Streaming and Remote Play.

We don't live in a world where cloud gaming services are an alternative to handheld consoles. At most they're an alternative to home consoles and gaming PCs to the relatively few people with a good enough internet connection.
 
How does xcloud replace a handheld gaming device?

The same way that a smartphone replaces a handheld gaming device for lots of people.
I don't have hard numbers, it's just anecdotal, but I know one person with a switch (the little smartphoneless daughter of a colleague), and a lot of people that use the smartphone to play.
 
nvidia always try to put old overpriced hardware to console (rsx in ps3, old tegra in switch) probably thats one of the reason it's no longer partner for microsoft or sony

I think that if nothing interesting shows up in the SDK during this year, this much is certain ...

It'll very likely disconfirm that Nintendo's successor will retain backwards compatibility ...
 
Thinking about it the xboy could play xbox one gen games natively and stream the series version of games if you wanted to play those games.

Also with the need to switch in and out of games quickly xbox has you covered with quick resume!

at this point in time it be silly to target the xbox one for its specs when there are modern apu designs. Using a future AMD cpu and gpu design should be able to keep up with the series s and of course with direct ml they can always upscale the image like dlss.
 
If you must target something, go directly for Series S to the point that any existing game can run on it without hassle.
If you must require a rewriting of the code, go to an architecture handheld friendlier, like arm.
In my opinion.
 
The same way that a smartphone replaces a handheld gaming device for lots of people.
I don't have hard numbers, it's just anecdotal, but I know one person with a switch (the little smartphoneless daughter of a colleague), and a lot of people that use the smartphone to play.

How many of those smartphone users are playing games on the cloud? They are probably playing games downloaded from Play / App Store. Your point does not confirm in any way your stance.
 
My point is about the return.
You have a huge market, and you can attack it now without any additional investment, and even reaching 2% of the share would be huge. More if you consider the basically 0 amount of investment.
On the other hand, you have a company that after 20 years can't break into a market where having the most powerful console is an advantage (while having the most powerful console), and a market where you can succeed only if you are nintendo.
They of course can make the best handled by just not using scraps from nvidia, they can force their studio to make games for it, and then what?
Possible, but not the most productive.
 
The new unannounced Switch model is still using the Mariko SoC.

As soon as the Mariko revision was released and tested I speculated that it would fit the bill for a Switch Pro model, assuming a Pro model would ever happen. The reduction in power consumption was significant, basically cut in half. It only made sense that this would allow for significantly higher clock speeds. Even overclocked OG Switch units have never shown issues with overheating, so if Nintendo were to improve the cooling system at all, even with Nintendo being very conservative, we could see a Pro model that pushes clock speeds 30-50% higher than the base Switch model. Mariko has a max GPU clock speed of 1.26Ghz, perhaps Nintendo goes with a big bump on the GPU while keep the increase to CPU clock speeds more modest.
 
My point is about the return.
You have a huge market, and you can attack it now without any additional investment, and even reaching 2% of the share would be huge. More if you consider the basically 0 amount of investment.
On the other hand, you have a company that after 20 years can't break into a market where having the most powerful console is an advantage (while having the most powerful console), and a market where you can succeed only if you are nintendo.
They of course can make the best handled by just not using scraps from nvidia, they can force their studio to make games for it, and then what?
Possible, but not the most productive.
if they made a portable console that ran the series S version of games natively then the investment and dev buy in wouldn't be that hard, you would already have all the software and because the only investment would be the hardware itself it wouldn't be that much overall investment. Also another reason that it wouldn't be a huge risk is because it wouldn't be replacing the series x/s, just complementing them, its not like you are properly changing console generations.


The thing is there is no one competing with Nintendo in the portable console market, there are certainly plenty of people who would like to play native versions of more demanding games in a portable form factor, look at how many people apparently buy doom on the switch, it would be going after that market.
 
Series S is about 100W, even with a node shrink it will be too power hungry by almost an order of magnitude, that converted in the imperial system is too much and 0.45inches.
To make a portable you need to invest in new hardware.
And I'm not saying that microsoft is not in the position to do it.
 
Series S is about 100W, even with a node shrink it will be too power hungry by almost an order of magnitude, that converted in the imperial system is too much and 0.45inches.
To make a portable you need to invest in new hardware.
And I'm not saying that microsoft is not in the position to do it.


I think they will likely just wait, AMD has been improving ryzen and RDNA by roughly ~50% perf/watt per year for the last couple of years, if they wait for three gens the same performance would be 12.5W, and you have to remember that one of those gens has just occurred with the release of RDNA 3 and the new ryzen cpus, so in reality they would be waiting for two gens. Maybe the time is right in 3-4 years time?
 
In about 4 years from now, we will be already talking about the damage caused by vaccines inoculated with 6G, and anything will be 5G even in the usa.
By that timeframe xcloud will be like watching youtube on the phone.
Switch is less than 9W at peak for the whole system, 12.5W only for the soc is a lot.
 
Microsoft wouldn't need to launch a mobile Series S to achieve a handheld console that does circles around what the Switch or a Switch Pro can do.

What would be the cost and power consumption of a 7nm SoC with 3 WGPs / 6 CUs at 1700MHz (same 1.3 TFLOPs) + 128bit LPDDR4X 2133MHz (same 68GB/s) + 8 Zen cores clocked at 1.75GHz (i.e. using transistors maximized for density) + 32MB eSRAM?
Van Gogh is probably way ahead of this already and it starts at 7.5W.

I bet Microsoft / AMD could make a portable XBoneS that consumes similar or even less power than an OG Switch.
 
In about 4 years from now, we will be already talking about the damage caused by vaccines inoculated with 6G, and anything will be 5G even in the usa.
By that timeframe xcloud will be like watching youtube on the phone.
Switch is less than 9W at peak for the whole system, 12.5W only for the soc is a lot.

I do think cloud gaming is the future but no way its gonna happen in 4 years time. Its not just about speed but datacaps. There is a reason Netflix and all the other streaming services offer downloads of shows and movies.
 
I understand that usa is a big market, and that for microsoft is almost the only one, but in other countries data cap start from a very minimum of 2GB/month to about 60GB/month for about 15$/month with voice included.
Considering a timeskip of 4 years, I'm decently optimistic about the comparison with netflix/youtube.
 
bet Microsoft / AMD could make a portable XBoneS that consumes similar or even less power than an OG Switch

Probably, but from concept to market that could take 2-3 years to get it out to market. Plus, who is going to provide the software to drive sales for the platform? We saw this with the Vita, very nice hardware without enough compelling content to drive sales for the platform. Regardless if its AMD, Nvidia or even Qualcomm, they all have the capability to design a SOC that fits into that sub 10w power requirement. If it were a contest, maybe Qualcomm would come out on top with the most performance per watt, but it wouldn't be a night and day difference. Regardless of which manufacture ultimately came out on top with the best performance, they would all be within a stones throw of each other. Similar to Xbox One vs PS4, PS4 is obviously the better performing hardware, but they sit comfortably in the same ballpark.

The old two Gamecube's duck taped together saying could come back around. Switch 2 could use a direct evolution of the Tegra X1 and double it all up. Configure it with a 128 bit memory bus, double the GPU cores to 512 clocked at 1.5 Ghz with eight A57 cores clocked at 1.7Ghz and splice in some tensor cores for DLSS and you have something that is comparable to a PS4 in portable form factor.
 
Back
Top