Contribute
Register

Solving NVIDIA Driver Install & Loading Problems

So they are. Odd that Nvidia described them as beta drivers in their press release.
They changed their PR about 2-3 hours later omitting the word "beta".
 
You have Inject/Nvidia/True. As indicated by the 0MB VRAM and that it names your card with the manufacturer. Change it to false.
I have tried: change it to false and reboot with nvda_dra=1.
It boot to black the same as before.
Thanks
 
I have tried: change it to false and reboot with nvda_dra=1.
It boot to black the same as before.
Thanks

If you have changed Inject/Nvidia to False and when you boot it still says 0MB VRAM; you aren't editing the right config.
 
I had my Hackintosh all working, dual-booting with Windows 10, and I used it for a while. In January I upgraded to a GTX 1060, and since they didn't work with Hackintosh I started using Windows exclusively (unfortunately).

When I saw that Pascal drivers for macOS came out I got really excited, and immediately tried to set my 1060 to work with my Hackintosh. I updated to 10.12.4, installed the drivers and rebooted. The Apple loading bar gets about halfway, then stops, and the display goes dark and my monitor goes to sleep. I have tried it with iMac14,1 and 14,2, as well as MacPro5,1, no luck, same problem. I updated Clover to no avail. I am now using iMac14,2 because another person helping me said that was the most compatible with my system.

If I use nv_disable=1 it boots fine (obviously a little choppy with the GUI though.

The boot flags I have now are: dart=0. Any thoughts would be greatly appreciated!

Parts:

  • Motherboard: Well I am not sure. I have a prebuilt Acer Predator G3-605, and have not been able to find out what motherboard I have for sure. If it helps the GTX 760 that came with the machine worked with it, and without the graphics card it works, so I don't think it is the motherboard.

  • CPU: Intel i7 4770k

  • GPU: EVGA GTX 1060 6GB

  • RAM: 8gb Storage: 256 GB SSD (Hackintosh partition with 128 GB, empty other ~128 GB).
 
Hi guys!

I need some help here. I recently added a Gigabyte 1060 3GB graphics card but I haven't been able to run it. My monitor gets no signal when i try to run it with the HDMI plugged into the Gigabyte GPU. It works find when its plugged into the integrated GPU though.

I have:
Updated to MacOS Sierra 10.12.4
Installed Nvidia Driver 378.05.05.05
Installed CUDA 8.0
Changed NvdiaWeb to true

When i boot up the Nvidia Web Driver is the default driver, no problem. Any help would be much appreciated. I have also uploaded my config.plist. Please help!
 

Attachments

  • config.plist
    8.1 KB · Views: 218
  • Screen Shot 2017-04-16 at 7.45.11 PM.png
    Screen Shot 2017-04-16 at 7.45.11 PM.png
    987.8 KB · Views: 155
Hi guys!

I need some help here. I recently added a Gigabyte 1060 3GB graphics card but I haven't been able to run it. My monitor gets no signal when i try to run it with the HDMI plugged into the Gigabyte GPU. It works find when its plugged into the integrated GPU though.

I have:
Updated to MacOS Sierra 10.12.4
Installed Nvidia Driver 378.05.05.05
Installed CUDA 8.0
Changed NvdiaWeb to true

When i boot up the Nvidia Web Driver is the default driver, no problem. Any help would be much appreciated. I have also uploaded my config.plist. Please help!
try removing ig-platform-id make sure bios is set for just discreet gpu and try AGDPfix
 
How would I manually enter the EmuVariableUEFI-64, because I used the given method twice and still no luck.
 
How would I manually enter the EmuVariableUEFI-64, because I used the given method twice and still no luck.

Assuming you installed the RC Scripts with the Clover installer, I actually had an issue with my ASUS motherboard's native NVRAM support not working correctly, so I used this approach but it wasn't restoring the saved nvram.plist at startup. I read thst some versions of Clover break this and others had success using earlier versions.

I worked around this last night with a very ugly hack to the boot script, which force-feeds a plaintext (vs. plist) copy of nvram as written by the (also hacked) shutdown script. Use of a plaintext dump was necessary since the nvram command doesn't read its own XML PLIST dumps. It now remembers things across reboots.

That's all the further I got before bed last night; I still have the black screen issue but I have yet to reapply the other fixes and I'm not allowed to sit for hours in my office playing with my toys today (Easter Sunday), so... ;-)
 
Back
Top