I believe MS has delayed the transition to 6nm because they are still unable to sell off the 7nm wafers they over-ordered.

Then as previously speculated.
The cost of chips at N4 is not high.

https://pastebin.com/V5nTeh4h

https://www.granitefirm.com/blog/us/2023/04/29/cost-of-chip-foundry/

This is a much better source then pastebin and a person tha actually works for the industry. Aside from that if it was nearly the same price why wouldn't Microsoft go with 4nm or 5nm? 

Is the answer cost?

Yes and it would also be why the reason Nintendo probably went 8nm cause it was that or 5nm which would be probably a lot  more expensive.

Well, I looked at the link you posted, but before I do, could you please look over the link I posted once.

There is a process node guess from 85 on the left and it says why tsmc 7nm and Samsung 8nm are not reasonable choices.

Then about the Xbox refresh.

What I am hearing is that the transition from 7nm to 6nm is officially recommended by tsmc and requires no redesign, so there is little cost to the transition.
Production costs will also be lower.

The Xbox refresh is about lowering costs and improving production capacity, not increasing performance.
That is probably why we chose 6nm.

And my point was that 8nm/7nm is a bad process node to go for if you are trying to minimize cost. Exemplified by both Sony and Microsoft switching away from it as fast as possible. 

Microsoft's goal with the refreshes is to reduce the size and power requirements, and therefore cost. Going 6nm makes sense for that purpose. 5nm would be overkill. 

Nintendo doesn't have an option to go 6nm as Nvidia isn't purchasing 6nm wafers. Their options are Samsung 8nm/7nm (what Nvidia purchases for Ampere) or TSMC 5/4nm (what Nvidia purchases for ADA Lovelace.)

Nintendo going with an 8nm Sammy node doesn't make sense from a cost perspective because it was a small disaster cost-wise for the other companies. 

It doesn't make sense from what we know of the spec either, for reasons mentioned multiple times in this thread and elsewhere on the internet.

8nm Sammy and 5/4nm TSMC are estimated to be nearly the same price after factoring transistor density and increased yields. 

Nobody said TSMC 6nm and TSMC 5/4nm, were (although as prices change and 5/4nm mature they might get closer.) 

Again, Nintendo can't just choose any node. They have to choose the nodes that Nvidia has purchased for their other chips. Nvidia isn't purchasing TSMC 6nm wafers. The options are Sammy 8nm/7nm, or TSMC 5/4nm. 

A (final cost of Sammy 8nm chip) ~= B ( final cost of TSMC 5nm chip) <=/=> B (final cost of TSMC 5nm chip) ~=C  (final cost of TSMC 6nm chip.)

I'm curious to see games on the switch 2. I watched a video of Rift on the deck and it looks like complete garbage. I'm hoping the switch 2 passes the deck up by a solid margin.   

Nintendo has used TSMC for decades now too.

The Wii U GPU was TSMC 40nm in 2012 ... 40nm TSMC production started in 2008, so Nintendo used a 4 year old process here.

The Switch GPU was TSMC 20nm in early 2017 ... 20nm TSMC production started in fall 2014, so Nintendo used a 2 1/2-3 year old process here.

If the Switch 2 is using TSMC 5nm for fall 2024 ... TSMC started production in fall 2020 (iPhone 12), so this would be a 4 year old process.

The Switch Lite/Mariko die shrunk to TSMC 16nm in 2019 ... TSMC started 16nm production in 2015 .... so this was a 4 year old process. 

This time line pretty much is in line with Nintendo's previous two consoles. Nothing outrageous here at all. Nintendo using a 4 year old TSMC node design is the same thing as the Wii U and actually more dated than the Switch 1 was. TSMC will probably want to keep that Nintendo business too, they are a fairly large customer that they've been working with for a long time. 

I'm certain it's gonna be 8nm but good luck. i read enough and use common sense to know that 8nm is 99% happening. what do you think the odds are for 5nm 50%?