Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Status
The first post of this thread is a WikiPost and can be edited by anyone with the appropiate permissions. Your edits will be public.
Can you start some version macOS (High Sierra) using the very same setup without flashing the card or changing the current vBIOS? Just put an SSD/HDD with a known to work installation in and wait for a long time unless you believe the macOS must have started. The best will be to enable remote login, screen sharing and file sharing in the Sharing system preferences pane in advance. This way you can check remotely if the came system up, but not the internal LCD.

Can you provide a download link for the Ubuntu image you have used.

As I told you the GRML Linux enables the text mode in the same way on all my systems and cards (having a Dell Bios or the BIOS from the first post)!

The interesting part is when you get a card with no suitable BIOS loaded at all or a HP card (just a stub file). Then neither GRML not Ubuntu will show a screen on the main LCD, at least this is my status as long as you prove me wrong.

One can setup any Linux distribution to provide the same functionality as the GRML one provided here. This is no church.

P.S.:

The 2011 system have an internal GPU (called iGPU) in the Intel CPU (named HD2000). This may be used by the Ubuntu, but this iGPU is not available in the Intel CPU with Late 2009 and Mide 2010 systems.
ubuntu-20.04.2-live-server-amd64.iso

Sorry I'm not interested in remote login or screen sharing since I'm not good at MacOS and now already am able to access the gpu via local console. (So no reason to try it yet and actually I was literally introduced to MacOS when I started to replace my iMac gpu after buying a few weeks ago. haha. Linux is much more comfortable to me.)

My 2 cents is Ubuntu and Debian based GRML are different packages and can be based on the different Kernel since GRML is for mostly light systems so It may not support various H/W drivers like Apple computers compared to Ubuntu. Also no need to make a custom package from Ubuntu since all you need is to xfer or copy the bios updater and one bios file or two to your iMac.

And if GRML boots up like Ubuntu and shows a login prompt when a new GPU is installed, that's cool since no need to try DHCP and set up remote SSH at all unless the gpu does not have bios!

But, as I commented earlier, my experience was my USB sticks running many distros didn't show any boot-up screen at all except Ubuntu Server.
 
Last edited:
Hello,

I am new here and quite a noob too. I am wondering If nickeys rom can be used as a solution for edp internal screen also on win laptop. I own dell precision m6800 with Nvidia k4100m and what I read till know It looks like there is no solution for dgpu with edp screen (only LVDS motherboards works).

Right now I am running Big Sur 11.3 accelerated by Intel HD4600. Nvidia dgpu works only via hdmi on external screen. If I disable optimus in BIOS I always get black screen.

Do you think If I can get working internal screen with this neckey's rom also on my laptop?

Thank you in advance!

Kind regards,
 
Hello,

I am new here and quite a noob too. I am wondering If nickeys rom can be used as a solution for edp internal screen also on win laptop. I own dell precision m6800 with Nvidia k4100m and what I read till know It looks like there is no solution for dgpu with edp screen (only LVDS motherboards works).

Right now I am running Big Sur 11.3 accelerated by Intel HD4600. Nvidia dgpu works only via hdmi on external screen. If I disable optimus in BIOS I always get black screen.

Do you think If I can get working internal screen with this neckey's rom also on my laptop?

Thank you in advance!

Kind regards,
Nobody has done this before to my knowledge. Save your vBIOS and give it a try. But please do not hijack this Apple iMac thread with a long off topic discussion.

You are welcome to use any technology published here and you can start a separate new thread about your Hackintosh.
 
ubuntu-20.04.2-live-server-amd64.iso

Sorry I'm not interested in remote login or screen sharing since I'm not good at MacOS and now already am able to access the gpu via local console. (So no reason to try it yet and actually I was literally introduced to MacOS when I started to replace my iMac gpu after buying a few weeks ago. haha. Linux is much more comfortable to me.)
We are going round in circles. You still not get the point!

You are just a lucky guy because you used only two Dell cards which are known to work in iMacs out of the box - although not offering a boot screen, brightness control and other features. Just read the original posters first published lines (it is at the bottom of the first post) and his subsequent posts from 2014. This guy bought a Dell Nvidia card and it worked in his iMac out of the box. Even without using Ubuntu, he just started his macOS version which I asked you to do and you refused. So far you did not really post any new information.

This afternoon I decided to end this discussion by using your Ubuntu Server with

i) an already flashed K5000M in a late 2009 - Ubuntu came up as expected showing the same selection as on your pictures

ii) an unflashed HP K610M in the very same machine. I disconnected the internal HDD and tried to boot from the external Ubuntu USB drive as before - screen stayed black, only two green LED showed up. A clear sign that the card itself is ok, but no usable BIOS available too drive an internal or extern screen.

One needs a working vBIOS to use the internal screen. In that corner case you can use possibly all Linux distributions you can grab to open text mode on the internal screen. Nothing new - familiar with this text mode using Emacs on the first Sun Workstation arrived at my university long, long ago.

Booting the GRML Linux flash utility allowed me to remotely access the system with the K610M (still black internal screen) and get the HP BIOS file (attached here). You see these cards have no BIOS image loaded. Please understand why we need remote access to overcome a not working internal screen on GPU change.

Code:
root@grml /lib/live/mount/medium/flash # ./nvflash_linux --save /root/K610M.rom
NVIDIA Firmware Update Utility (Version 5.414.0)
Simplified Version For OEM Only
Adapter: Quadro K610M         (10DE,12B9,103C,197B) H:--:NRM  S:00,B:01,D:00,F:00
Identifying EEPROM...
EEPROM ID (EF,3012) : WBond W25X20A 2.7-3.6V 2048Kx1S, page
Reading adapter firmware image...
IFR Data Size         : 1016 bytes
IFR CRC32             : 26EC83DB
IFR Image Size        : 1024 bytes
IFR Image CRC32       : A8AC4411
IFR Subsystem ID      : 103C-197B
Image Size            : 0 bytes
Version               : Unavailable (Invalid) (IFR-only Image)
~CRC32                : 00000000
Image Hash            : N/A
Subsystem ID          : 103C-197B
Hierarchy ID          : None
Chip SKU              : N/A
Project               : N/A
CDP                   : N/A
UEFI Support          : No
UEFI Version          : N/A
UEFI Variant Id       : N/A ( Unknown )
UEFI Signer(s)        : Unsigned
InfoROM Version       : N/A
InfoROM Backup Exist  : NO
License Placeholder   : Absent
GPU Mode              : N/A
Saving of image completed.
root@grml /lib/live/mount/medium/flash #

Code:
root@grml /lib/live/mount/medium/flash # ./nvflash_linux -6 Quadro_Beta1.2/K610M_EG2.rom
NVIDIA Firmware Update Utility (Version 5.414.0)
Simplified Version For OEM Only
Checking for matches between display adapter(s) and image(s)...
Adapter: Quadro K610M         (10DE,12B9,103C,197B) H:--:NRM  S:00,B:01,D:00,F:00
NOTE: Only IFR present:
NOTE: Skipping Board ID and Device ID checks.
NOTE: EEPROM does not contain board ID, skipping board ID check.
WARNING: Firmware image PCI Subsystem ID (106B.1026)
  does not match adapter PCI Subsystem ID (103C.197B).
Please press 'y' to confirm override of PCI Subsystem ID's:  y
Overriding PCI subsystem ID mismatch
Current      - Version:Unavailable (Invalid) ID:0000:0000:103C:197B
Replace with - Version:DE.AD.00.BE.EF ID:10DE:12B9:106B:1026
               Quadro K610M by Nick[D]vB (Normal Board)
Update display adapter firmware?
Press 'y' to confirm (any other key to abort):  y
Identifying EEPROM...
EEPROM ID (EF,3012) : WBond W25X20A 2.7-3.6V 2048Kx1S, page
WARNING: Adapter does not have a preservation table,
unable to preserve board settings.
Storing updated firmware image...
................................................................
Verifying update...
Update successful.
Firmware image has been updated from version Unavailable (Invalid) to DE.AD.00.BE.EF.
A reboot is required for the update to take effect.
./nvflash_linux -6 Quadro_Beta1.2/K610M_EG2.rom  47.02s user 2.67s system 80% cpu 1:01.46 total
root@grml /lib/live/mount/medium/flash #

After flashing GRML, Ubuntu and macOS could be booted using the internal screen!

P.S.: At lease one positive result, the K610M is working properly!
 

Attachments

  • K610M.rom.zip
    918 bytes · Views: 80
Last edited:
We are going round in circles. You still not get the point!

You are just a lucky guy because you used only two Dell cards which are known to work in iMacs out of the box - although not offering a boot screen, brightness control and other features. Just read the original posters first published lines (it is at the bottom of the first post) and his subsequent posts from 2014. This guy bought a Dell Nvidia card and it worked in his iMac out of the box. Even without using Ubuntu, he just started his macOS version which I asked you to do and you refused. So far you did not really post any new information.

This afternoon I decided to end this discussion by using your Ubuntu Server with

i) an already flashed K5000M in a late 2009 - Ubuntu came up as expected showing the same selection as on your pictures

ii) an unflashed HP K610M in the very same machine. I disconnected the internal HDD and tried to boot from the external Ubuntu USB drive as before - screen stayed black, only two green LED showed up. A clear sign that the card itself is ok, but no usable BIOS available too drive an internal or extern screen.

One needs a working vBIOS to use the internal screen. In that corner case you can use possibly all Linux distributions you can grab to open text mode on the internal screen. Nothing new - familiar with this text mode using Emacs on the first Sun Workstation arrived at my university long, long ago.

Booting the GRML Linux flash utility allowed me to remotely access the system with the K610M (still black internal screen) and get the HP BIOS file (attached here). You see these cards have no BIOS image loaded. Please understand why we need remote access to overcome a not working internal screen on GPU change.

Code:
root@grml /lib/live/mount/medium/flash # ./nvflash_linux --save /root/K610M.rom
NVIDIA Firmware Update Utility (Version 5.414.0)
Simplified Version For OEM Only
Adapter: Quadro K610M         (10DE,12B9,103C,197B) H:--:NRM  S:00,B:01,D:00,F:00
Identifying EEPROM...
EEPROM ID (EF,3012) : WBond W25X20A 2.7-3.6V 2048Kx1S, page
Reading adapter firmware image...
IFR Data Size         : 1016 bytes
IFR CRC32             : 26EC83DB
IFR Image Size        : 1024 bytes
IFR Image CRC32       : A8AC4411
IFR Subsystem ID      : 103C-197B
Image Size            : 0 bytes
Version               : Unavailable (Invalid) (IFR-only Image)
~CRC32                : 00000000
Image Hash            : N/A
Subsystem ID          : 103C-197B
Hierarchy ID          : None
Chip SKU              : N/A
Project               : N/A
CDP                   : N/A
UEFI Support          : No
UEFI Version          : N/A
UEFI Variant Id       : N/A ( Unknown )
UEFI Signer(s)        : Unsigned
InfoROM Version       : N/A
InfoROM Backup Exist  : NO
License Placeholder   : Absent
GPU Mode              : N/A
Saving of image completed.
root@grml /lib/live/mount/medium/flash #

Code:
root@grml /lib/live/mount/medium/flash # ./nvflash_linux -6 Quadro_Beta1.2/K610M_EG2.rom
NVIDIA Firmware Update Utility (Version 5.414.0)
Simplified Version For OEM Only
Checking for matches between display adapter(s) and image(s)...
Adapter: Quadro K610M         (10DE,12B9,103C,197B) H:--:NRM  S:00,B:01,D:00,F:00
NOTE: Only IFR present:
NOTE: Skipping Board ID and Device ID checks.
NOTE: EEPROM does not contain board ID, skipping board ID check.
WARNING: Firmware image PCI Subsystem ID (106B.1026)
  does not match adapter PCI Subsystem ID (103C.197B).
Please press 'y' to confirm override of PCI Subsystem ID's:  y
Overriding PCI subsystem ID mismatch
Current      - Version:Unavailable (Invalid) ID:0000:0000:103C:197B
Replace with - Version:DE.AD.00.BE.EF ID:10DE:12B9:106B:1026
               Quadro K610M by Nick[D]vB (Normal Board)
Update display adapter firmware?
Press 'y' to confirm (any other key to abort):  y
Identifying EEPROM...
EEPROM ID (EF,3012) : WBond W25X20A 2.7-3.6V 2048Kx1S, page
WARNING: Adapter does not have a preservation table,
unable to preserve board settings.
Storing updated firmware image...
................................................................
Verifying update...
Update successful.
Firmware image has been updated from version Unavailable (Invalid) to DE.AD.00.BE.EF.
A reboot is required for the update to take effect.
./nvflash_linux -6 Quadro_Beta1.2/K610M_EG2.rom  47.02s user 2.67s system 80% cpu 1:01.46 total
root@grml /lib/live/mount/medium/flash #

After flashing GRML, Ubuntu and macOS could be booted using the internal screen!

P.S.: At lease one positive result, the K610M is working properly!
I gotcha!

What I wanted to say was no need to ignore or underestimate updating vbios via the local console at all. The guide tells you two major ways to update the gpu which are 1. remote access via Linux/Windows and 2. using CHA341A and its programmer tool. So my point is we can use this local consol update as a major method to reimage your vbios and think remote ssh for headless gpu as a second option. I even bought cheap mxm adapter to update two GPUs and failed to recognize them in windows. If I knew Ubuntu can do, I wouldn't buy at all.

Booting up to MacOS using unmodified gpu does not provide a way to update your vbios so it's less meaningful than booting up to Linux and updating it, no?

So I mean all we can do is trying booting up from iMac itself first and trying another method like remote ssh later.
 
Last edited:
I gotcha!

What I wanted to say was no need to ignore or underestimate updating vbios via the local console at all. The guide tells you two major ways to update the gpu which are 1. remote access via Linux/Windows and 2. using CHA341A and its programmer tool. So my point is we can use this local consol update as a major method to reimage your vbios and think remote ssh for headless gpu as a second option. I even bought cheap mxm adapter to update two GPUs and failed to recognize them from windows. If I knew Ubuntu can do, I wouldn't buy at all.

Booting up to MacOS using unmodified gpu does not provide a way to update your vbios so it's less meaningful than booting up to Linux and updating it, no?

So I mean all we can do is try booting up from iMac itself first and trying another method like remote ssh later.
It does not end and it does not bring any new information at all. Instead I have to explain step by step what happens- This is of no help at all. We all know that we can (in rare cases) use the console to update the BIOS. Most users will be completely in the dark if you feed the impression, that this is the normal case. It is not!

Using the (GRML or other) Linux you can even flash the BIOS (remotely) without putting the internal LCD in. In 5% of the case you have a dead GPU and in 30% of the cases the Late 2009 showing the internal black screen. So betting on a Nvidia GPU or AMD GPU with a working BIOS is ok, but as I told you you can use the GRLM Linux then to do the BIOS update in front of the internal screen. The remote Linux or Windowas method works all the time while your method works only in special corner cases. Be happy to find one from time to time but do not claim you have found another or new BIOS installation method overcoming the remote access method entirely. You do not oversee all the pitfalls of process.

And your complete Ansatz is even worse: Betting on a working VBIOS would need to install the LCD, check this out (and because you have no boot selection you need to disconnect all internal disks upfront). After flashing you will have to open the iMac again, reconnect internal disks and reinstall - risking only more damage to the system. We had in the last week only three damaged logic boards after that operation.

Keep the process simple and clean.

This is my last message about that. Booting into macOS would have been another proof that the BIOS is running from the start (what I tried to explain no several times to you).

Honstestly I meet rarely a guy so obsessed with a bad idea like you.
 
Last edited:
I gotcha!

What I wanted to say was no need to ignore or underestimate updating vbios via the local console at all. The guide tells you two major ways to update the gpu which are 1. remote access via Linux/Windows and 2. using CHA341A and its programmer tool. So my point is we can use this local consol update as a major method to reimage your vbios and think remote ssh for headless gpu as a second option. I even bought cheap mxm adapter to update two GPUs and failed to recognize them from windows. If I knew Ubuntu can do, I wouldn't buy at all.

Booting up to MacOS using unmodified gpu does not provide a way to update your vbios so it's less meaningful than booting up to Linux and updating it, no?

So I mean all we can do is try booting up from iMac itself first and trying another method like remote ssh later.
If you want to cover all possible options, then there are more than 2 options available.
1. Install to a PC, with PCIe 1x to MXM adapter. Windows (or MS DOS) GUI or Linux text command line
2. Install to a PC with native MXM slot (the laptop where those MXM card were taken from). Windows (or MS DOS) GUI or Linux text command line
3. CH341a clip to the Vbios chip. CH341a ROM programer. Windows PC only.
4. Install to the iMac MXM slot, Xanderon Linux USB, then SSH from another machine. (no monitor attached)
5. Install to the iMac MXM slot, OS=Ubuntu server (your option), and hope that the internal LCD show image.

Options 1 to 3 require running Windows, which many Mac users hate or are not prepared for.
Option 4. is more commonly applicable to Mac only users, and all necessary guidelines have been provided in this thread.
Option 5 is a hit or miss, if the internal screen doesn't show any image, one would have to go back with option 4 anyway.
 
Honstestly I meet rarely a guy so obsessed with a bad idea like you.
Thank you. haha!

This is my last message about that. Booting into macOS would have been another proof that the BIOS is running from the start (what I tried to explain no several times to you).

I also have told you a few times that I couldn't get any MacOS boot from those two gpus. All I had to do was replacing the gpu and re-assembling everything back including LCD and waiting. Nothing happened.
 
Last edited:
If you want to cover all possible options, then there are more than 2 options available.
1. Install to a PC, with PCIe 1x to MXM adapter. Windows (or MS DOS) GUI or Linux text command line
2. Install to a PC with native MXM slot (the laptop where those MXM card were taken from). Windows (or MS DOS) GUI or Linux text command line
3. CH341a clip to the Vbios chip. CH341a ROM programer. Windows PC only.
4. Install to the iMac MXM slot, Xanderon Linux USB, then SSH from another machine. (no monitor attached)
5. Install to the iMac MXM slot, OS=Ubuntu server (your option), and hope that the internal LCD show image.

Options 1 to 3 require running Windows, which many Mac users hate or are not prepared for.
Option 4. is more commonly applicable to Mac only users, and all necessary guidelines have been provided in this thread.
Option 5 is a hit or miss, if the internal screen doesn't show any image, one would have to go back with option 4 anyway.
Thank you for your comments.
Btw going from option 5 to option 4 is easy. You just have to reimage your usb stick with another iso image. You can download two isos in case and trying both. That's all.
 
  • Like
Reactions: Nguyen Duc Hieu
If you want to cover all possible options, then there are more than 2 options available.
1. Install to a PC, with PCIe 1x to MXM adapter. Windows (or MS DOS) GUI or Linux text command line
2. Install to a PC with native MXM slot (the laptop where those MXM card were taken from). Windows (or MS DOS) GUI or Linux text command line
3. CH341a clip to the Vbios chip. CH341a ROM programer. Windows PC only.
4. Install to the iMac MXM slot, Xanderon Linux USB, then SSH from another machine. (no monitor attached)
5. Install to the iMac MXM slot, OS=Ubuntu server (your option), and hope that the internal LCD show image.

Options 1 to 3 require running Windows, which many Mac users hate or are not prepared for.
Option 4. is more commonly applicable to Mac only users, and all necessary guidelines have been provided in this thread.
Option 5 is a hit or miss, if the internal screen doesn't show any image, one would have to go back with option 4 anyway.
1. difficult, the HP one I owned hat only place for MXM-A cards
3. would not work with any (of this) NVIDIA card unless the BIOS chip has been changed.
5. does not work on Late 2009 and on HP cards and possibly on a lot of NVIDIA cards, too.

We are rewriting everything what have been found out during the last years just because this guy is reluctant to read it himself. Stop this.
 
Last edited:
  • Like
Reactions: Nguyen Duc Hieu
After days of waiting for my K1100M to arrive, I was finally able to install it on my iMac. I was already running Ubuntu anyway so I just downloaded the files needed to flash the K1100M. I installed the card and flashed it. Installed macOS Big Sur 11.3.1 with OCLP on the installer flash drive. After that, it was pretty much working right off the bat. I updated macOS via OTA to 11.4 and no hiccups so far. I've also installed Windows 10 alongside it too and it's been great!

I'll probably upgrade my Wi-Fi card next to get AirDrop working. Haha!

Thanks to everyone who made this possible!
 

Attachments

  • Screen Shot 2021-05-26 at 2.41.07 PM.png
    Screen Shot 2021-05-26 at 2.41.07 PM.png
    374.6 KB · Views: 142
1. difficult, the HP one I owned hat only place for MXM-A cards
3. would not work with any (of this) NVIDIA card unless the BIOS chip has been changed.
5. does not work on Late 2009 and on HP cards and possible on a lot of NVIDIA cards, too.

We are rewriting everything what have been found out during the last years just because this guy is reluctant to read it himself. Stop this.

I have no doubt that option 4 with the Xanderos Linux USB is the most convenient method and have widest appliable range.
 
I updated macOS via OTA to 11.4 and no hiccups so far.

I'll probably upgrade my Wi-Fi card next to get AirDrop working. Haha!

Thanks to everyone who made this possible!
I updated to 11.4 too and so far so good. I'm also interested in Wifi update but not good at soldering so dunno if I can do it good.
 
I have no doubt that option 4 with the Xanderos Linux USB is the most convenient method and have widest appliable range.
The main point is: Some one will have to do the user support. Having 20 half working methods only confuses the people.

We had this with the seven different ways to install Big Sur and even now we have some die hards proposing Patched Sur (it's journey ends with the last full installer in summer) or the micropatcher (same thing).

As I wrote, this is no church and unless these ideas get no traction I am fine with it. Otherwise we know who will have to explain the does and don't of half working method on a daily basis.
 
The main point is: Some one will have to do the user support. Having 20 half working methods only confuses the people.

I simply had to try a different method since GRML was half baked soution not getting DHCP IP address at all and I couldn't find the reason w/o screen. (I still don't know the reason yet) My laptop booting from the same USB stick worked perfectly getting IP address instantly.
my method could update vbios easily in the end.
Why ignoring when sharing my experience so the others can try? Nothing to lose.
 
ubuntu-20.04.2-live-server-amd64.iso

Sorry I'm not interested in remote login or screen sharing since I'm not good at MacOS and now already am able to access the gpu via local console. (So no reason to try it yet and actually I was literally introduced to MacOS when I started to replace my iMac gpu after buying a few weeks ago. haha. Linux is much more comfortable to me.)

My 2 cents is Ubuntu and Debian based GRML are different packages and can be based on the different Kernel since GRML is for mostly light systems so It may not support various H/W drivers like Apple computers compared to Ubuntu. Also no need to make a custom package from Ubuntu since all you need is to xfer or copy the bios updater and one bios file or two to your iMac.

And if GRML boots up like Ubuntu and shows a login prompt when a new GPU is installed, that's cool since no need to try DHCP and set up remote SSH at all unless the gpu does not have bios!

But, as I commented earlier, my experience was my USB sticks running many distros didn't show any boot-up screen at all except Ubuntu Server.
Why would I want to download and run Ubuntu server @ 1.13gb when I can download and run a persistent version of GRML @ 360mb with most of the current vbios roms pre loaded by just copying and pasting some folders to a USB.

I understand you had some problems and I'm happy you overcame that with your solution I personally would have tried to figure out why the @xanderon USB wasn't working.

I'm actually hoping I or someone else will come up with a persistent tiny core usb boot disk with SSH over wifi at under 100mb.
 
Last edited:
I simply had to try a different method since GRML was half baked soution not getting DHCP IP address at all and I couldn't find the reason w/o screen. (I still don't know the reason yet) My laptop booting from the same USB stick worked perfectly getting IP address instantly.
my method could update vbios easily in the end.
Why ignoring when sharing my experience so the others can try? Nothing to lose.
GRML is NOT a half baked solution. Your situation is EXTREMELY rare. I have been following this thread since @xanderon gave us the GRML solution. There was ONE other case where the DHCP wasn't working. However, there are hundreds and hundreds of users that have used GRML flawlessly. Your "method" will only work in rare circumstances. As pointed out by @Ausdauersportler, your method WILL NOT WORK when there is no vBios present, which is more frequent than you would think.

So, it's great you found a solution to YOUR problem that coincidentally worked BUT, please, let's not clog this thread with a "shot in the dark" method.
 
GRML is NOT a half baked solution. Your situation is EXTREMELY rare. I have been following this thread since @xanderon gave us the GRML solution. There was ONE other case where the DHCP wasn't working. However, there are hundreds and hundreds of users that have used GRML flawlessly. Your "method" will only work in rare circumstances. As pointed out by @Ausdauersportler, your method WILL NOT WORK when there is no vBios present, which is more frequent than you would think.

So, it's great you found a solution to YOUR problem that coincidentally worked BUT, please, let's not clog this thread with a "shot in the dark" method.
Indeed, it's really exaggerated because the GRML solution is such a big contribution to this community. But it's not a 100% working solution like my case. I'm guessing it's something similar to this case.
Unless someone could fix this issue, it's not perfect yet.

I was almost giving up this upgrade if ssh via Wifi was not feasible but it's really easy trying my workaround when you encounter DHCP problem.
The best solution would be repackaging grml or whatever distro by adding support of wifi at iMac (as The_Croupier said) or enhancing dhcp/staic ip assignment and ssh access by some experts. Personally, I do not want to check my router to see which IP my iMac got whenever I do this job. Supporting static IP by a certain network wizard or by a config file will be cool.
 
Last edited:
Indeed, it's really exaggerated because the GRML solution is such a big contribution to this community. But it's not a 100% working solution like my case. I'm guessing it's something similar to this case.
Unless someone could fix this issue, it's not perfect yet.

I was almost giving up this upgrade if ssh via Wifi was not feasible but it's really easy trying my workaround when you encounter DHCP problem.
The best solution would be repackaging grml or whatever distro by adding support of wifi at iMac (as The_Croupier said) or enhancing dhcp/staic ip assignment and ssh access by some experts. Personally, I do not want to check my router to see which IP my iMac got whenever I do this job. Supporting static IP by a certain network wizard or by a config file will be cool.

Actually, for advance users like you, you don't even need to remote SSH or using a console, even with DHCP problems.
There is a solution which PC guys call "blind flash method". It is described in details on techpowerup.com.
Using the Xanderon USB, or your Ubuntu USB, you just need to create an autoexec.bat file in the root folder of Linux USB which included the flash command and correct path to the ROM files, as well as correct deviceID (the graphic card). When you boot from the USB, the flashing process will be done automatically and after that, it will shutdown the machine (add a shutdown command to the autoexec.bat). It's just a little scary when you don't have any visual confirmation that the flashing process was successful or not.

And I did connect an external monitor to my iMacs when booting with Xanderon's Linux USB, hoping that I didn't have to SSH to it from another machine.
No images were displayed in both iMac 2009 or iMac 2010. I had to SSH anyway.
 
Last edited:
Actually, for advance users like you, you don't even need to remote SSH or using a console, even with DHCP problems.
There is a solution which PC guys call "blind flash method". It is described in details on techpowerup.com.
Using the Xanderon USB, or your Ubuntu USB, you just need to create an autoexec.bat file in the root folder of Linux USB which included the flash command and correct path to the ROM files, as well as correct deviceID (the graphic card). When you boot from the USB, the flashing process will be done automatically and after that, it will shutdown the machine (add a shutdown command to the autoexec.bat). It's just a little scary when you don't have any visual confirmation that the flashing process was successful or not.
That kind of automation is still good but if I can use CH341A and can grab the tiny eeprom with the clip correctly, it's still safer to save the original rom and update via the tool manually I guess.
btw, when we say auto-update, it's a shell script at the end of init process in the linux instead of running windows .bat file.
 
Last edited:
After a lot of research on here and the help of the very kind members on here. I narrowed it down to couple a cards, I only use the iMac for work and the odd bits of photo and light video editing, no gaming.

Just purchased this GPU WX4130 2GB, the seller stated it's Genuine pulled from a Dell laptop. Will be flashing using the USB Linux method. I will be running it on my iMac 11,1 27inch Late 2009 i7 2.8GHz 16GB Ram 1TB SSD with OpenCore and Big Sur. Hopefully it all goes to plan. Just waiting on the 1mm Copper shim and Kapton tape. Will let everyone know how it goes.

IMG_1902.jpeg


IMG_1903.jpeg
 
  • Like
Reactions: globetribe
** NVIDIA Geforce GTX880M Mac Edition ROM **
** NVIDIA Geforce GTX870M Mac Edition ROM **
** NVIDIA Geforce GTX860M Mac Edition ROM **

Genuine Native Boot Screen & Brightness Control


View attachment 942200

Pre-installation Requirements:
- iMac12,2 (27-inch Mid 2011) Mac-942B59F58194171B (Tested machine)
- BootROM: 87.0.0.0.0. Please use High Sierra to update your BootROM to latest version.
- One of the following GPUs:

NVIDIA GeForce GTX 880M
N15E-GX-A2, MXM-B (3.0)
8GB VRAM

NVIDIA GeForce GTX 870M
N15E-GT-A2, MXM-B (3.0)
non-functional cards/original_ROMs:
80.04.EF.00.A2 (MSI)
80.04.F7.00.0C (MSI)

NVIDIA GeForce GTX 860M
Kepler (GK104)
N15P-GX-B-A2
Base ROMs:
80.04.F5.00.06
80.04.E8.00.0D
80.04.E9.00.03
Maxwell (GM107) N15P-GX-A1? - non-function with MacOS; cannot trust sources with GPU Variant
Base ROMs:
82.07.24.00.19
82.07.34.00.03
82.07.34.00.08
82.07.49.00.05

The GTX880M is the first 8GB VRAM card to work in our machines! I now have as much vram as system ram. I'm happy to report that macOS sees all the ram. It is boosting fully even at base clock speeds. When I purchased this card, the rivets were very long and kept too much of the GPU away from the heatsink surface. It was thermal throttling early as it sensed an overheat situation. When I removed the rivets, and used screws to secure it, I was able to get a much tighter seal and the card boosted properly.

Tested on a 2011 iMac using High Sierra 10.13.6. Please feel free to test on other MacOS versions, I will update this post as necessary with success/failures.

  • These ROMs do not require a 3rd party bootloader like OpenCore.
  • They will require a modification of the base AppleIntelPanelA/ApplePanels/F10Ta007 brightness stepping.
  • The frame-buffer depth issues remains for now, and can be temporarily corrected by entering a sleep cycle.

As previous, these roms should bring back:

⦁ Genuine native brightness control
⦁ Genuine ‘gray’ early-boot screen (stage 1 & 2 progress bar)
⦁ Genuine macOS bootloader compliance

Post-installation Requirements:
Brightness Control Stepping Mod:
-Turn computer on, hold down Command(⌘)-R
-Choose Utilities > Terminal
-Enter:csrutil disable
-MacOS Catalina: requires you to make root writeable: sudo mount -uw /
-Reboot
-Download and open 'Kext Utility v2.6.6'

-Navigate to S/L/E (System/Library/Extensions)
-Copy "AppleBacklight.kext" to Desktop
-Edit: AppleBacklight.kext/Contents/Info.plist
-Scroll down to: IOKitPersonalities > AppleIntelPanelA > ApplePanels
-There you find several Apple LCD profiles.
-For the iMac 2011 27" machine locate:
Code:
<key>F10Ta007</key>
<data>
ABEABgALABQAHAAnADMAPwBOAFwAZwBzAIEAkQClAL8A2wD/
</data>
-Change the <data> section to:
ABEAAgA3AF8AigCzAOsBJAFnAakB1AIJAlQCogL4A00DlgRpBGk=

-Drag your modded kext into Kext Utility, allow it correct permissions
-"Applebacklight.kext.bak" folder will be created
-Reboot

The above data pattern will allow for a wider span of steppings for the brightness control and utilizes more of the capacity of the HD3000. If you have a different machine, your panel ID can be found by going to System Preferences > Displays > Color > Open Profile > mmod

Caveats post-install/Bugs:
-16bit resolution glitch of UGA_DRAW_PROTOCOL - temporary solution: activate a sleep cycle and return, this should now clear the issue.

**UPDATE**
9-4-2020: please note the 870M_6GB_UGA.rom is experimental and in testing phases for those interested and have the card.
5-3-2021: I've removed 870M_Clevo_BR.rom because it is not functional and replaced it with GTX870M_V1.rom, appreciate @jimac2011 for testing.

"insanely great!"
-Steve Jobs

View attachment 940773 View attachment 940774
Hi, @nikey22

Are you able to check if this GPU can be supported like the other GTX 870M? It has slightly different version than the others commented here before.

NVIDIA Firmware Update Utility (Version 5.414.0)
Simplified Version For OEM Only
Adapter: GeForce GTX 870M (10DE,1199,1462,1106) H:--:NRM S:00,B:01,D:00,F:00

Identifying EEPROM...
EEPROM ID (C2,2012) : MX MX25L2005 2.7-3.6V 2048Kx1S, page
Reading adapter firmware image...
IFR Data Size : 1084 bytes
IFR CRC32 : ED3783C7
IFR Image Size : 1536 bytes
IFR Image CRC32 : 72B987C6
IFR Subsystem ID : 1462-1106
Image Size : 95232 bytes
Version : 80.04.F7.00.06
~CRC32 : B068A7E6
Image Hash : 688E5AA7A3EF639B72852C3ADD349A00
OEM String : NVIDIA
Vendor Name : NVIDIA Corporation
Product Name : GK104 Board - 20530003
Product Revision : Chip Rev
Device Name(s) : GeForce GTX 870M
Board ID : E137
PCI ID : 10DE-1199
Subsystem ID : 1462-1106
Hierarchy ID : Normal Board
Chip SKU : 650-0
Project : 2053-0003
CDP : N/A
Build Date : 03/21/14
Modification Date : 04/09/14
UEFI Support : No
UEFI Version : N/A
UEFI Variant Id : N/A ( Unknown )
UEFI Signer(s) : Unsigned
InfoROM Version : N/A
InfoROM Backup Exist : NO
License Placeholder : Absent
GPU Mode : N/A
Sign-On Message : GK104 P2053 SKU 003 VGA BIOS (HWDIAG).
 
Last edited:
Sierra and High Sierra Package Installer

( iMac 2009-2011 AGC, FakeSCM, Lilu, WhateverGreen, AppleBacklightFixup installer)

Attached you will find the a kernel extension package installer currently working on Sierra 10.12 and later. SIP needs to be disabled. Boot into recovery, start Terminal app, enter csrutil disable and reboot directly into your macOS and install the package.

It will patch the current installed AppleGraphicsControl (AGC) as described in this post addressing the black screen issue and install the AppleBacklightFixup needed for ++ type Nvidia cards side by side with the @vit9696 Lilu and Whatevergreen. As an option you can have the FakeSMC extensions get installed.

The NVIDIA ++ cards are this running with a BIOS from @nikey22 and marked in the table on post #1 as type ++.

After starting the Apple package installer (double click) you can choose the destination and select the packages to be installed according to your wishes. For this reason I separated the complete sets in several subsets. You can use the installer to patch AGC (this will be done in any case), and install addionally BacklightFixup, FakeSMC, and WhateverGreen/Lilu. You have to select at least one kernel extension to be installed. Only the FakeSMC is really optional.

FakesSMC will be needed b the HW Monitor app to provide full details on all sensors available on your system.

What will this package installer do:
  1. Patching AppleGraphicsControl AGC (all iMacs, Nvidia card users, only)
  2. Nvidia BacklightFixup (all iMacs, NVIDIA ++ cards users, only!
  3. FakeSMC Package (all iMacs, all graphics cards)
  4. vit9696 WhateverGreen and Lilu (all iMacs, all graphics cards)
Supported hardware:

iMac 11,x (Late 2009 and Mid 2010) with NVIDIA Kepler GPU
iMac 12,x (Mid 2011) with NVIDIA Kepler GPU


Supported MacOS:

macOS Sierra 10.12.1 up to macOS Catalina 10.15.7
do not use with macOS Big Sur!


Recipe:

Disable SIP (Mojave and Catalina will have this done), download package, double click, enter password if asked for, select disk and your extensions to be installed, finish installation and reboot.

You may select only a single of the packages below! The AGC patch will be applied in any case.

View attachment 1666152


Attached you will find some selections screen shots. You can use the installer to select a single package only.

Have fun and please report back!

Notes:
12/07/2020: Late Nikolaus gift, new package installer and deinstall package
Hi,
I had a black screen and have connected an external display via TB/HDMI, works fine. But I wanted my old screen back to live. Installing the patch delievers me two black screens. Tested all the stuff noted but no success anymore iMac 2011, i5 3,4 , K4100M flashed version.
Any idea how to reactivate the external connection. Thanks, Greetings
 
Hi,
I had a black screen and have connected an external display via TB/HDMI, works fine. But I wanted my old screen back to live. Installing the patch delievers me two black screens. Tested all the stuff noted but no success anymore iMac 2011, i5 3,4 , K4100M flashed version.
Any idea how to reactivate the external connection. Thanks, Greetings
Yes, you could have used to read the first post and just tried to use alt/option on boot to force the internal screen to come up.

Try a PRAM reset. If this does not help reinstall High Sierra from the scratch and next time please tell the complete story (aka firmware, macOS version) and so on. Do a firmware upgrade in case you did not. There is no i5@3,4GHz :)

EDIT: I described the alt/option trick just to make sure you do not suffer a hardware problem which we could not solve by patching.
 
Last edited:
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.