|
|
|
|
|
|
by striking
3352 days ago
|
|
|
> their Open Source driver for Linux is so much crap it was entirely rejected by the kernel team No, amdgpu was not rejected by the kernel team. A particular implementation of the driver was rejected because it implemented an abstraction layer, and that would make it nearly impossible for kernel devs to maintain. > drawing too much power from the mobo and damaging it If you could point to an example of this happening, I'd appreciate it. My knowledge of the situation is that some models of the RX 480 can run slightly out of spec, pulling a little too much power from the motherboard. Any motherboard I've heard of could withstand that. And if you really care you can enable an option in the driver that causes it to run strictly in PCI spec. I'm not an AMD shill, I just think you've misrepresented some of the issues at hand. AMD make mistakes, for sure. But not every mistake is as crippling as you've implied. |
|
|
Just look in the AMD own official forums for threads created before they made driver patches, for example in one thread a guy put a photo of his molten blockchain mining rig, and then there was several pages of people calling him a nVidia shill, and noone helping.
The handling of the incidents were so bad I stopped visiting AMD forums entirely, it was just pure hostility to anyone with any problem, even unrelated problems.