I wanted that for so long. But there was a rumor around of nvidea pushing for Nintendo to upgrade cuz they wanted to push their new architecture and nintendo was like no thanks were good. I honestly think Nintendo realised that their brand does not depend on power like the other 2. 

I doubt it. The most likely thing I see hapening is they have active cooling on the dock. That way they can overclock it a lot more in dock mode. That might actually be good as i think the switch even docked dosent get to the highest clock the tegra can reach. 

And? Who cares. I debated the points presented.

You are missing my point however.. .That flops is essentially useless unless you understand it's purpose... And 99% of forum respondents don't. It's a buzzword.

Even before the "GPU" existed, video cards tended to be highly scalable in terms of functional units.

I disagree. 1536 is very low-end in this day and age... nVidia doesn't even make an RTX GPU with that amount of functional units.

I think there will be a hit to clockrates, no doubt. Mobile always does.

You might as well sell us two consoles. A fixed home console and a portable based on the same hardware.

I think Nintendos approach with scalable clockrates is the best approach honestly.

Correct... The current Switch piggybacks off USB which isn't fast enough or low-latency enough to handle the data required to be transmitted between both devices.

But if a Switch 2 doesn't use the USB protocol and uses something like... PCI-Express, then it would become technically possible.

But I think the general consensus today is that multi-GPU approaches are a waste of time and effort... It doesn't scale linearly and it increases developer burden and thus the likelihood of erratic performance and bugs. (This coming from someone who used to run with 4x Radeon in crossfire at one point.)

USB 4 Gen 3x1, Gen 2x2, Gen 2x1 is the same bandwidth as USB 3's variants. That is... 1.2 to 2.4GB/s. Just different encoding.

USB 4 Gen 3x2 doubles things to 4.8GB/s. Which is still not fast enough.
Keep in mind it's got to transmit more than just display or audio data.

USB can take 125ms or more for a transaction.
30fps needs 32.2ms. 16.7ms for 60fps. - Starting to see an issue?

Nowhere near it infact.

The Switch tops out at 768Mhz... And people have taken the Switch and overclocked it to 921Mhz. And it can definitely scale above even that as Maxwell Tegra will happily run at 1.26Ghz when allowed.

On chip power, I'm not up to speed (nor anywhere near an expert) on how powerful Nvidia chips are going to be in the next 2-3 years. But when subtracting the Apple tax, it doesn't seem out of the ballpark that Switch could have something more powerful than the basic M1 caliber SoC which is:
CPU - 8 cores at 3.2 GHz
GPU - 8 cores
RAM - 8 GB
This combined with a 256 GB solid state drive
The Mac mini costs 649 USD, the Macbook Air costs 849 USD. Surely Nintendo and Nvidia could bring out something greater than this power for much cheaper in 2023-2026.

We already have a good idea about which chip Nintendo will use. Yesterday nVidia announced the cheapest Orin chip, the NX. 6 core CPU, 1024 GPU cores, Ampere, 8GB LPDDR5. It's very cheap (1/4 the cost of the full Orin chip). It would be a good Steam Deck competitor, and with carefully written software exceed it. I expect that's what we will get.

Personally I just want Nintendo to grab an RTX 3060 type product and stick that in a $400 console, desktop only. It would be easy to do since no DVD drive or hard drive is required, making it even cheaper to make than the PS5 digital. Let people add their own hard drives and storage like the Switch. It would be a true hybrid, since you could play all your Switch games and all the PS5 level games. The Switch people call it a hybrid, but really all it does is play mobile level graphical games. So it is a mobile only system, very limiting. See how that works? A desktop system is actually not limiting. Sure you can't carry it around, but it can play twice as many games. A DESKTOP CAN PLAY MOBILE AND DESKTOP GAMES.

The Radeon 7870 in the PS4 was more expensive than the RTX 3060, and yet the PS4 had it, plus a lot of expensive add-ons (optical drive, hard drive, and CPU cores). RTX 3060+6 ARM cores, done. Voila. Cheap and fast desktop home console.

But if we can't have that, we'll get the Orin NX.

I really feel it's this https://developer.nvidia.com/embedded/jetson-orin Jetson NX

I think you are dead on man. If I had to guess, it would be the Jetson NX 16GB.

Saying TFLOPS don't mean anything is like saying clock frequency doesn't mean anything, or core count, or really any one variable of a computer. It's just a ridiculous assertion. It isn't everything, it isn't even most things, but it gives you an idea of the hardware's capabilities. This is why, generation after generation, FLOPS increase substantially. Because they are correlated with gaming capability, it's just not 1:1. 

Also, we do know FP16 performance given the above information. We don't know TOPS, but we can assume it'll be less than 200 (probably 100) based on what we know about Orin. We can attempt to calculate the TMU's based on what we know about Orin though. Orin also has a maximum of 16 SM's. However, the GPU that is closest to the calculated TFLOPS that I provided has 14 SM's. Given that the RTX-20 line has 4 TMU's per SM, we are looking at a maximum of 64 Gtexels, with a likelihood of 56 Gtexels/s if we stick with the 1Ghz frequency. The thing is it wont be that frequency. We don't know about ROP units, but I don't think I've seen any GPU with higher ROP count than TMU count, so 56 Gpixels/s is really the maximum here. The Series S has a pixel rate of 50GPixels/s and a texture rate of 125 GTexels/s. Unless Nintendo has no customization of the card (e.g. doesn't reduce ROPs) and doesn't decrease the frequency (reduction will absolutely happen) then we know it isn't going to match the Series S GPU. It can get close on the pixel rate side, but it'll be behind on the texture rate. 

But, again, a lot of this is guess work and the only substantive evidence we have is TFLOPS (which freebs2 brought to my attention may not actually be substantive). We have information on Orin that isn't about TFLOPS, which is what I'm using a lot of my guesswork about. So the above is what I'm basing my knowledge on with it not being equivalent to Series S in raw power. And, sure, raw power isn't everything, but there is nothing that we can factually discuss that isn't related to raw power in this time. 

Hmm interesting. When I was doing some digging I thought that it was confirmed, but it does seem like it is an assumption and the source of the assumption isn't quite known. Still, I don't think that number is really that crazy, though I do agree it would be the upper limit. I don't think we would essentially be getting an off shelf T234. Too much power draw even if you were to cut the frequency. 

I'm not sure about the lower clock frequencies. Yes, it's definitely true that it will have more power per watt, but that doesn't necessarily mean the power necessary for the chip to function at these frequencies wont be high. Lastly, I would assume that the Orin chip will be bulkier than the X1, which may necessitate a smaller battery for a conservative form factor. 

Honestly, you're right though. There are just so many variables, so even though the leak tells us it's Orin, it's not really narrowing down what the Switch 2 will be capable of my as much as everyone would hope. 

Yeah that's the one I was talking about. I don't think we'll get the 16GB model though. I mean, the Deck has 16GB so Nintendo could do it, but you could technically port PS5 level games to an 8GB system with worse textures and a lower resolution. I'd like the 16GB though. I think textures are great looking, and don't lower FPS performance. If Valve can do it, why not Nintendo. LPDDR5 must be dropping in price a lot recently.

8GB or 16GB, the important part is getting 4x the memory bandwidth you can finally handle 1080p / 60 fps.