Guru3D NVMe Thermal Test - the heatsink vs. performance

- after years telling people that 9/10 motherboards had awful design with the M.2 above the gpu or on it's side getting all the hot air
(the only good one was the MSI TRX40 Creator who had two m.2 at the bottom of the motherboard and 1 vertically on the right of the ddr slots, all 3 away from heat sources and in places that had better airflow)
- that the 5.0 M.2 from the latest asus Z690 extreme was the most stupid thing ever and was UNUSABLE (because not only does it kill your gpu bandwith, currently it divides it by 4x not 2x...4x !) but it means their stupid idiotic giant heatsink 1mm above the gpu backplate is actively HEATING your M.2 "who whould have guessed that placing a heat exchanger 1mm above a 80-100°C plate would not cool it...woaw I'm shocked !"
(that's what you get when you claim to give useful info to consumers but keep using test benches because it makes YOUR job easier...but then all your measurements are trash because an open test bench is not a real computer aka a closed case)

I'm afraid that's too late for the next AMD lineup because I noticed on the announcement they also looked like they had heatsinks almost touching the gpu, so another fake slot that you really shouldn't use, not like you want it anyway...see my rants on why pcie 5.0 is worst than 4.0 (you gain nothing but only lose more bandwith when the x16 is split between the pcie slots)
maybe that's the reason why AMD went dual chipset...I hope so
when they say : your pcie slot is

x16 x8 x4

it's not : x of 5.0 speed
it's : of whatever pcie speed your peripheral is
so when your gpu pcie slot goes from 16x 5.0 (52Tb/s) down to x8 it's not x8 5.0 (26Tb/s) it's x8 4.0 at best which is 1/4 speed or x16 3.0 (13Tb/s) and yes you lose performance with high end videocards