Contribute
Register

How to build your own iMac Pro [Successful Build/Extended Guide]

Status
Not open for further replies.
Awesome! Thank you so much, I am going to test this one. This should work just fine for High Sierra if I just use the usb patches from your original post? I can't run Mojave as I need to have cuda power with Octane Render. Thanks!

Disable USB port limit patches and use 15-port USB kext instead of fully inplemented XHC USB kext.
 
WOW, I wish I would have found your post on the Deluxe II! I should have NOT bought this board! I still can't see if anyone has gotten hot-swap working. Is the onboard TB working properly? I am going to do my tests too.

TB works, but apparently no hotplug under macOS. TTR USB-C is only implemented with respective devices connected at boot.

However note that I do not own a Deluxe II myself. All results and conclusions have been derived based on the input and feedback provided by the three guys mentioned above.

I don‘t know if your other post in the X299 10.14 thread is still necessary, as you are anyway on 10.13.6, which could just create unnecessary discussion and confusion in any case.
 
Last edited:
TB works, but apparently no hotplug under macOS. TTR USB-C is only implemented with respective devices connected at boot.

I don‘t know if your other post in the X299 10.14 thread is still necessary, as you are anyway on 10.13.6, which could create unnecessary discussion and confusion in any case.
Ok, thank you for the clarification! I will be continuing to test today and be reporting back. I still can't get a proper shutdown, any suggestions on that? I did use your EFI folder.
 
Ok, thank you for the clarification! I will be continuing to test today and be reporting back. I still can't get a proper shutdown, any suggestions on that? I did use your EFI folder.

Proper shutdown? I told you that you need to adopt the SSDTs for your additional PCIe hardware and slot configuration! E.g., the EFI-Folder contains a SSDT-X299-RX580.aml, which is definitely not suited in your case! You should also remove the fully implemented USB kext, unzip the 15-port USB kext and disable the port limit patches in the config.plist, etc., etc. etc. ... Study this small guideline for a proper SSDT adaptation and else take profit of available SSDT approaches provided in the X299 SSDT Github library.

Take your time before unnecessarily posting issues. I guess, I already helped you a lot for now. You should get the rest of things done by yourself. If really necessary, others might be also able to help you with remaining issues, if there are any at all. Just ask @AsEvil, @applemacosxGOD or @Maywitt for help if really necessary. But please here in this thread as you are on 10.13.6 and please not in the 10.14 thread!

BTW.. you know that CUDA drivers can mess up your system? And concerning your multi-multi-multi Nvidia GPU approach I am sceptic and pessimistic anyway. I abandoned Nvidia long time ago and with 10.14, Nvidia is history now anyway on state-of-the-art Hackintosh systems where Metal also substitutes CUDA.

Good luck, my friend!

KGP
 
Last edited:
Proper shutdown? I told you that you need to adopt the SSDTs for your additional PCIe hardware and slot configuration! E.g., the EFI-Folder contains a SSDT-X299-RX580.aml, which is definitely not suited in your case! You should also remove the fully implemented USB kext, unzip the 15-port USB kext and disable the port limit patches in the config.plist, etc., etc. etc. ...

Take your time before unnecessarily posting issues. I guess, I already helped you a lot for now. You should get the rest of things done by yourself. If really necessary, others might be also able to help your with remaining issues, if there are any at all.

Good luck,

KGP
I guess I should have mentioned I did do that. I swapped the RX580.aml for the Nvidia.aml and I have nothing in the PCI-e slots as of now, just the 1080ti in slot one. I also used the 15 port usb kext. I also disabled the mojave usb kexts. Here is the efi folder attached. I followed all your steps so far. I really do appreciate your help, and taking your time with all of us non Jedi`s!
 
I guess I should have mentioned I did do that. I swapped the RX580.aml for the Nvidia.aml and I have nothing in the PCI-e slots as of now, just the 1080ti in slot one. I also used the 15 port usb kext. I also disabled the mojave usb kexts. Here is the efi folder attached. I followed all your steps so far. I really do appreciate your help, and taking your time with all of us non Jedi`s!
 

Attachments

  • EFI.zip
    36.7 MB · Views: 173
I guess I should have mentioned I did do that. I swapped the RX580.aml for the Nvidia.aml and I have nothing in the PCI-e slots as of now, just the 1080ti in slot one. I also used the 15 port usb kext. I also disabled the mojave usb kexts. Here is the efi folder attached. I followed all your steps so far. I really do appreciate your help, and taking your time with all of us non Jedi`s!

My friend, you cannot just copy and paste the Nvida SSDT from the Github Library.

Compare PCI path and PCI replacements in SSDT-X299-RX580.aml for slot-1 of the Deluxe II with those in the Nvidia SSDT for slot-1 on the Deluxe! They are different!

In my previous post I already provided you with a link to a guideline how to properly adopt a SSDT based on the original ACPI information (IOREG).

After properly adapting/modifying the Nvidia SSDT for slot-1 on your Deluxe II system, confirm the proper SSDT implementation again by means of IOREG and Section "PCI" of Apple's system report!

I am answering your questions now already for one hour although I do not provide user specific support for now. It really makes no sense that you try to do things in a rush. You should carefully read the guidelines, understand the underlaying basics and try to get things properly done in consequence. Just take the necessary time for doing it! And you will need to do it anyway if you really aim at adding multi-multi-multi Nvida GPUs and eGPUs to a 3x x16 PCIe slot board.. ;)

It is midnight here.. Off to bed... All the best for your endeavour.

KGP
 
Last edited:
My friend, you cannot just copy and paste the Nvida SSDT from the Github Library.

Compare PCI path and PCI replacements in SSDT-X299-RX580.aml for slot-1 of the Deluxe II with those in the Nvidia SSDT for slot-1 on the Deluxe! They are different!

In my previous post I already provided you with a link to a guideline how to properly adopt a SSDT based on the original ACPI information (IOREG).

After properly adapting/modifying the Nvidia SSDT for slot-1 on your Deluxe II system, confirm the proper SSDT implementation again by means of IOREG and Section "PCI" of Apple's system report!

I am answering your questions now already for one hour although I do not provide user specific support for now. It really makes no sense that you try to do things in a rush. You should carefully read the guidelines, understand the underlaying basics and try to get things properly done in consequence. Just take the necessary time for doing it!

It is midnight here.. Off to bed... All the best for your endeavour.

KGP


Thank you as always I will take your advice and be working away. Goodnight KGP! Thanks again!
 
  • Like
Reactions: kgp
Thank you as always I will take your advice and be working away. Goodnight KGP! Thanks again!

One last comment before jumping to bed... Ever thought on implementing Vega 64/Radeon VII GPUs and Radeon VII eGPUs instead of your multi Nvidia GPU approach? There is yet no native implementation for Radeon VII GPUs/eGPUs by Apple and one still has to see if the Radeon VII can be used at all on X299/X99 systems OoB after a proper native macOS implementation. But in my personal opinion, that's really something you should think about. Nvidia under macOS will be just painful for your endeavour. And after all, I don't know if it would not be more clever to change to the ASUS WS X299 Sage 10G (7x x16 PCIe slots) + GC-TTR PCIe adapter in this case. If I would have the necessary budget, I would do the change to the ASUS WS X299 Sage 10G even in my case.

Good night,

KGP
 
One last comment before jumping to bed... Ever thought on implementing Vega 64/Radeon VII GPUs and Radeon VII eGPUs instead of your multi Nvidia GPU approach? There is yet no native implementation for Radeon VII GPUs/eGPUs by Apple and one still has to see if the Radeon VII can be used at all on X299/X99 systems OoB after a proper native macOS implementation. But in my personal opinion, that's really something you should think about. Nvidia under macOS will be just painful for your endeavour. And after all, I don't know if it would not be more clever to change to the ASUS WS X299 Sage 10G (7x x16 PCIe slots) + GC-TTR PCIe adapter in this case.

Good night,

KGP

Yes, I really have thought about that. The main problem a lot of us have is that all the 3D rendering apps such as Octane and Redshift ONLY uses Nvidia Cuda enabled GPUs for render power, so as of now we are stuck with 10.13.6 and Nvidia. I wish I could run an ATI, I have in a Mojave build and it was perfect! I think the best thing now is to run one 1080ti in 10.13.6 and buy a server to run the rest of the GPUs and do a network render solution over 10G.
 
Status
Not open for further replies.
Back
Top