We can see nothing, because there is nothing to show by Nintendo.

I can't even find Tegra X2 demos online :/

Also.. you shouldn't take 4GB RAM as face value. It actually has nothing to say on performance and features.

Oh, ok. Thanks.

Tegra X2 would at least pull the memory bandwidth closer to X1, 50 GB/s vs 68 GB/s. X1 has the much faster esram to help out yet main memory would be more comparable.

It's just speculation anyway. Tegra x2 is 0.75 tflops, better than x1 at 0.5 tflops. Funny the Tegra X2 now also has Int8 compute, 3 dltops!

Drive PX2 can do 6 tflops / 20 dltops! (at 80 watts, 2x tegra x2 + 2x Pascal MXM GPU)
http://wccftech.com/nvidia-tegra-parker-soc-hot-chips/

Anyway forget close to xb1 performance, the WiiU uses 34 watts total while gaming. The Switch also needs to power the screen and won't even use half that, doubt it can even power the x2 on battery mode. Maybe x2 has a low power mode for mobile?

What I can give Nintendo credit for is the very low power consumption. Where back in the day the PS3 and Xbox 360 used 160-180 Watts, the Wii only did 30 Watts and the Wii U like you said 34 Watts.

The PS4 is currently around 135 Watts and the Xbox One about 115 watts.

What power consumption the NS will have is only to be rumored. Both in handheld mode and home console mode.

The WiiU Gamepad uses about 1.8 watts (3 hours on a 1500 mah battery) Mobile phones have up to 3300 mah batteries, Vita 2200 mah with external battery pack rated at 5000 mah. It's not much to work with, Switch running at 6 watts will still give it only 3 hours with a 5000 mah battery.

At this point I doubt how much faster it will be than the WiiU. WiiU has 0.35 tflops, 12.8Gb/s memory + 70Gb/s edram. Tegra X1 doubles the main memory bandwidth but doesn't have edram and will only be 40% faster. WiiU to Switch would be comparable to X1 to PS4 difference. It's still impressive to pack 1.4x WiiU into a handheld, but I'll wait for DF before deciding if Zelda is worth upgrading to Switch for.

I wouldn't take much value in the flops. They don't represent the feature level the GPU/CPU will work with and what visuals will be possible in the end. It's like back in the day where they used BITS to describe the power of a console (8-bit, 16-bit, 32-bit, 64-bit and so on). But, you know.. compare N64 to Atari Jaguar 64 ... and do the math. Today we talk about FLOPS which is kinda the same not saying much about the whole performance of a console. Speaking about the battery the gamepad has.. that was kind of a joke, because they deliberately place a very small one into the gamepad.

The top one (1.5 A) is the default battery which is being place into the gamepad. The larger one (2,55A) gives you 2-3 additional hours of gameplay. And it really shows the differents.

Maybe the NS does the same? :/

I don't know the current power consumption of the NS, nor does I know the battery size. 5mah woud be pretty big. It's all a mystery as to what the games will look on the go and what part the docking station really plays.

This news is most awesome because each teraflop will make every game more fun.

True, however it's rather that mobile gpu's seem to have less grunt per rated gflop in real life. The Galaxy S7 Edge should have easily outperformed the wiiU at 512 gflops fp32, except for the catching on fire part :) And that's a 3600 mAH battery, 2.4x of the standard WiiU gamepad.

The Switch looks pretty big, 5,000 mAH battery is not impossible. The controllers have there own batteries as well, perhaps adding to the total battery life when attached. Anyway there's always the option to have an external battery, should be cheaper now after the Pokemon Go rush on them.

please be true.
memory bandwidth seems to be the main bottleneck, and x1 has only 25Gb/s.nowhere close o xbox one.
Double that, and its somewhat fine.
For the rest, just keep it cheap and somewhat good battery life.

Hopefully it is the X2. Now I just need someone to leak their first year lineup.