Question PCH PCIE Express Configuration

[hadynm]

This question is in reference to the answer from a previous thread

https://forums.tomshardware.com/thr...processor-pci-express-config-all-3-0.2726388/

I am looking at the ASUS TUF H310-Plus Gaming motherboard and was wondering.

The Spec sheet says

"With x4 PCI Express® 2.0 bandwidth, M.2 supports up to 20Gbps data-transfer speeds. It's the perfect choice for an operating system or application drive, making your PC and apps work as fast as possible."

Does this mean that if you have a x16 Graphics Card on it's own, in a all 16 lanes in use by the first slot configuration, and then plug in the Optane Memory chip, that the 16 lanes in use by the graphics card will be reduced to 8, and 4 of those reduced lanes will be redistributed to the M.2 slot, and four of the lanes will be wasted? And if so why design it this way at all?

Secondly, I noticed that the description says "PCI Express® 2.0 bandwidth" Your answer describes 16 3.0 lanes but doesn't mention 2.0 Lanes? Are the 2.0 lanes separate from the 16 3.0 Lanes?

 

[hadynm]

also, if using the optane does reduce the primary graphics card slots access to the cpu pcie lanes to 8, can buying a more expensive graphics card mitigate the performance loss you would sustain from those 8 lost lanes or would it be futile or cost inefficient to try?

I guess what I'm getting at is

Is it better to leave the pcie bus dedicated to graphics and try to get the fastest communications between the Memory and PCH storage or better to reduce graphics performance and retask the PCIE bus?

Or put SImpler... if you're needing high graphics performance, dont optane, and if you're doing lots of low graphic high memory to data work, do optane. Would these assumptions be about right?

Has anyone benchmarked low end graphics cards on x16 against high end graphics cards on x8? Is there a thread here on that?

I found this https://www.gamersnexus.net/guides/2488-pci-e-3-x8-vs-x16-performance-impact-on-gpus

Based on this article would it be fair to say that most modern graphics cards will perform equally well if they are the only card in the system and in the primary PCI-E x16 slot and the slot is operating on only 8 of the 16 available lanes than if they were using all 16 lanes?

If so then shouldn't practically all motherboard designs try to make use of the remaining x8 available in the bus and, in the instance of the motherboard I described, have 2 PCIE x4 M.2 Slots instead of 1?

What am I missing here? Why would the manufacturer release a model that only makes use of 12 of the 16 PCIE lanes? Only answer I can think of is not enough room on the PCB