Sorry for my poor English.
New drivers arrive every month and I install them but none fix the bug.
Before, that was easy: I had to resume the old version of April 2021 and everything was repaired.
The problem today is that this new driver (1.1.22), as bad as others, has removed this option and the squares are there even with the April version.
It must have broken something.
Now I have blue squares everywhere even on the Garmin!!!
Does anyone on this planet got a solution?
Is there someone competent at AMD who can answer?
Direct 12X unusable: too many crashes!
First … I share your frustration that driver version rollback is always such a nightmare.
Second … I can not help as I also do not have a solution
(maybe a little off topic:)
but … I would like to know more about your DX12 trouble.
I did switch to DX12beta because it does a least fix the tile chaos … even when then I get the new (related) “missing airport tile overlay” bug. But this is visually way less disturbing.
I do get (very) frequent CTD even when loading into the new departure location. This is somewhat random as repeated attempts for exactly the same airport and departure spot (ramp, gate) sometimes work … and sometimes crash.
So I suspect there is a DX12 specific “initial 3D asset loading CTD” bug … perhaps due to a race condition somewhere in the code or the driver.
But besides that I have not seen CTD crashes during the rest of the flight.
So … any further info on your DX12 CTD issue would be interesting (e.g. perhaps a pointer to a DX12 forum thread related to your issues)
Don’t know if this helps, I get this exact problem with my 6800xt - but only when I use it with my iMac through an egpu. I don’t get the problem when I use the same card installed within my pc (not in the egpu case). Haven’t tried dx12 yet. Maybe this suggests it’s a bandwidth issue as the egpu case restricts the card to 4 PCIe lanes (I think) rather than 16
So when you use it in your iMac means … you run Bootcamp … boot regular Windows … and use MSFS there? … or are you using VirtualBox or that like to run W10+MSFS and then give it direct PCI access to the eGPU?
It could be possible that “transfer speed” plays a role. But given that older AMD drivers did not have the issue on the same hardware it feels more like a driver bug … now if that is triggered by some race condition … then “faster transfer” might make it less likely to occur.
Sadly I suspect that AMD does not really care about MSFS … and so nobody seems to be really trying to debug it (I would gladly be proven incorrect on this one)
… and it feels like a pretty nasty “low level” bug … so it might take some real effort to locate and fix it.
Personally I have given up the hope that we will ever see an AMD driver fix for the DX11 issue … that is why I switched to DX12 (even when it seems to produce more CTDs during asset load stages)
My apologies for this poor English.
Thank you for these clever thoughts.
I think, unfortunately, Nenenui is absolutely right about AMD’s attitude…
Fortunately, (I did a little manipulation), I managed to restore my old driver (April 2021) which works well. and everything is now working fine on DX11. Bye, bye squares…
DX12 is actually too “buggy”
I’ve havenn’t really flown for a while (because of the bug) but first time I noticed it was a few updates ago at Heathrow. There was corruption around the gate I started at but nowhere else and the rest of the flight was OK. I then tried VFR from a few small airfields and they were all bug free.
That Entebbe was by far the worst I’ve experienced it.
Has anyone tried the very latest ‘optional’ AMD drivers ?
I’m still using the recommended.