I flew too close to the 1TF sun and my Switch wings melted.

[mutantsushi]

curl-6 said:

I'm not saying run it at full speed undocked or anything, that's obviously not viable, but 30% of a Tegra X1 is really, really slow. 
And since most devs are likely going to use undocked mode as the base level for graphics and simply up-res it when docked, 
the games aren't really going to look much better than Wii U this way.

Yes, that is the reality it seems.

JRPGfan said:

mutantsushi said:

I think it does have bearing on Switch in revealing consequences of cheaping out on performance strategy.
...Switch also commits to obsolete 2D 20nm process when future fab improvements (10nm, 7nm) will be 3D FinFET.
Rather than getting on board same fab node bandwagon everybody else is, even if aiming for smaller APU at lower clock,
Nintendo uses last-gen fab node which makes their design less amenable to future cost reductions from fab improvements.
It is aiming far behind the curve, and thus will always miss the largest % of cost improvements which the mainstream gains.

But cheap as f***.
Nvidia has tons of these 3 year old chips they made laying around, they cant sell
...If they launch it at 199$ they re going to be off to a great start. Handheld users will be willing to get it early/day1 too at that price.

Is there a specific basis for idea this is not new production, or just supposition?
Regardless, I would say it is falling into the same trap, because if it is moderately succesful, they will need new production at some point.
As you say, they won't have problems selling the initial batch between diehard fans and those solely comparing it to 3DS.
But basing the design on "garage sale" unsustainable price rather than sustainable production does not bode well for long term viability.
Not only do they give up the easy 30% performance:watt gains a modern 14nm FinFET design would allow,
a design based on 2D layout would need total rework (not just shrink) to use future 3D FinFET processes (14nm, 10nm, 7nm).
If they were so cheap to choose this in the beginning why spend later for less benefit?  
A modern 14nm FinFET design is not financially prohibitive now, and would enable easiest shrink path in future.
And importantly, while perhaps still not competitive with MS/Sony, it would allow solid step up in graphics from Wii U.
The chosen approach can't really do that, portable being equivalent to Wii U, and docked having power to output that at 1080p.
Obviously, we have some Wii U games already at 1080p, and there will be some graphics gains there,
but I'm talking opportunity to demonstrate clear cut "next generation" of Nintendo graphics, which isn't here IMHO.