4 ms·

"Apple hardware from 2012-2015 was really great" You didn't have your MacBook Pro spontaneously reboot because of the NVIDIA problem? I don't know if it was t

"Apple hardware from 2012-2015 was really great" You didn't have your MacBook Pro spontaneously reboot because of the NVIDIA problem? I don't know if it was the driver or the hardware but it wasn't fun.

yingw787 8y ago
Hasn't happened to me. If it did, I might have switched completely already :-) In fact that was why I returned the XPS 15, the OEM version of Ubuntu 18.04 (which did have correct sleep/hibernate behavior and worked well with the OEM battery) couldn't detect the NVIDIA Quadro P2000 card at all, which I wanted to run CUDA with. That persisted after about three factory resets, so I was like kthxbai.
- m_mueller 8y ago
  This sort of thing is why I've resigned to just use WSL (after switching from Mac). Works well enough.
seba_dos1 8y ago
...or troubles with SATA cables seemingly designed to break in the most subtle ways.
LeftTurnSignal 8y ago
If that's the same issue I had, it was a defective nvidia chip. I went through 2 logicboards and I know several of my friends who had 2012 MBPs went through a few too. We used to be an Apple shop, and we had a lot come in due to that issue.
zeristor 8y ago
I'm still using my MacBook Pro from 2010, and I'm working on ways to mitigate the NVidia card rebooting at its worse it was 120 times a month, but I've managed to get it down to about 10 times a month. Although IDEs seem to cause frequent reboots, but using Jupyter is far more stable.
- aspenmayer 8y ago
  I found this[0] fix when I last researched this problem for a client. It seemed to substantially reduce the problem. The only drawback is the need to re-run it after MacOS updates, but I don't think this functionality has been the target of any updates from Apple for some time anyway. [0] https://github.com/julian-poidevin/MBPMid2010_GPUFix https://github.com/julian-poidevin/MBPMid2010_GPUFix
BoorishBears 8y ago
The NVIDIA problems hit laptops from many manufacturers. I remember buying an HP TX 1000 with a faulty NVIDIA chip to try and reseat the GPU via heat (it didn't work but it was a cheap experiment) https://www.youtube.com/watch?v=JZ1i78kmb0M https://www.youtube.com/watch?v=JZ1i78kmb0M

yingw787 8y ago

Hasn't happened to me. If it did, I might have switched completely already :-) In fact that was why I returned the XPS 15, the OEM version of Ubuntu 18.04 (which did have correct sleep/hibernate behavior and worked well with the OEM battery) couldn't detect the NVIDIA Quadro P2000 card at all, which I wanted to run CUDA with. That persisted after about three factory resets, so I was like kthxbai.

m_mueller 8y ago
This sort of thing is why I've resigned to just use WSL (after switching from Mac). Works well enough.

seba_dos1 8y ago

...or troubles with SATA cables seemingly designed to break in the most subtle ways.

LeftTurnSignal 8y ago

If that's the same issue I had, it was a defective nvidia chip. I went through 2 logicboards and I know several of my friends who had 2012 MBPs went through a few too. We used to be an Apple shop, and we had a lot come in due to that issue.

zeristor 8y ago

I'm still using my MacBook Pro from 2010, and I'm working on ways to mitigate the NVidia card rebooting at its worse it was 120 times a month, but I've managed to get it down to about 10 times a month. Although IDEs seem to cause frequent reboots, but using Jupyter is far more stable.

aspenmayer 8y ago
I found this[0] fix when I last researched this problem for a client. It seemed to substantially reduce the problem. The only drawback is the need to re-run it after MacOS updates, but I don't think this functionality has been the target of any updates from Apple for some time anyway. [0] https://github.com/julian-poidevin/MBPMid2010_GPUFix https://github.com/julian-poidevin/MBPMid2010_GPUFix

BoorishBears 8y ago

The NVIDIA problems hit laptops from many manufacturers. I remember buying an HP TX 1000 with a faulty NVIDIA chip to try and reseat the GPU via heat (it didn't work but it was a cheap experiment) https://www.youtube.com/watch?v=JZ1i78kmb0M https://www.youtube.com/watch?v=JZ1i78kmb0M