Hello everyone! I have an imac 27 2011, i7 2600, GTX 780 4Gb. BigSur is installed. Maybe someone has encountered such a problem: when I run a test in "Unigine Valley", after some time my mac turns off, or goes into a reboot. I had the same problem on the last imac 27 2011, i5 2400, GTX 770 3Gb. But then I did not attach any importance to this. Maybe I'm missing something in the setup? In general, the mac works well and stably... but there is a suspicion that the video card does not work at its full capacity. Thanks!
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
iM 4,1-14,4 2011 iMac Graphics Card Upgrade
- Thread starter MichaelDT
- WikiPost WikiPost
- Start date
- Sort by reaction score
- Status
No big deal, it was just the GPU overheated under stress. That happens when you stress a 122W TDP GPU.
Strange, I have a cooling system for 3 pipes. Instead of HDD, there is an SSD. 1920x1080, quality Medium, stereo 3D disabled, Anti-aliasing Off and the test failed not once.
Your card was purchased brand new with warranty, or some piece of old parts salvaged from an old gaming laptop, having gone through all kind of torturing over the last decade?
Interesting. How did you find that?
Also: Would it be possible to overcome the ‘no POST scenario’ just with a proper GPU VBIOS? Or would the iMac’s BootROM (and possibly SMC firmware) also need to be modified? (If my questions sound like nonsense, please remember that I really don’t understand/know iMac’s pre-POST hardware checks…
)With the AMD cards you'll always get 2 diagnostics LED as it will never recognise the Screen on Boot, I had same issue and I thought the same my card was DOA and was going to give up. (Note and my green Dell WX4130 2gb is working flawlessly to date)
You need to do a PRAM reset first with the Linux USB Flash Stick plugged in and Ethernet cable connected otherwise you'll never be able to do the flash.
Follow this guide for the AMD flash steps https://forums.macrumors.com/thread....1596614/page-545?post=29723850#post-29723850
This should then work. Note you'll always get black screen on boot as Native Boot Screen is not supported with these cards. Not a big issue for most as soon as OSX / OCPL boots you'll see the screen.
If a card doesn't let the iMac boot, even when the VBIOS EEPROM is blank, then there's nothing that software could change.Interesting. How did you find that?
Also: Would it be possible to overcome the ‘no POST scenario’ just with a proper GPU VBIOS? Or would the iMac’s BootROM (and possibly SMC firmware) also need to be modified? (If my questions sound like nonsense, please remember that I really don’t understand/know iMac’s pre-POST hardware checks…
A GOP VBIOS might not make much difference as well, as our old iMac EFI doesn't use that GOP driver - but I don't know how much initialization is performed when loading the driver.
FirmwareInfo: usBootUpVDDCIVoltage (apple560X)Interesting. How did you find that?
Also: Would it be possible to overcome the ‘no POST scenario’ just with a proper GPU VBIOS? Or would the iMac’s BootROM (and possibly SMC firmware) also need to be modified? (If my questions sound like nonsense, please remember that I really don’t understand/know iMac’s pre-POST hardware checks…
And possibly Diffs at GpioPinLut.
Yestuday i sucessfully switch Wx4150 Pegatron Vbios from desktop to Mobile with adding this to GPIO_Pin_LUT
Here is that additional pin (LowPowerState, HighPowerState i think) result
As a clarification:
The Apple EFI driver seems to perform at least some hardware initialization without referring to the ATOM BIOS. I don't know of any investigations yet that reveal whether it has any references at all to the ATOM BIOS - I might just try creating a VBIOS without any ATOM tables at all in order to test.
The macOS drivers however, and the Windows drivers even more, do use the ATOM tables.
When reference clock reaches mxm PCIBus, the gpu core starts initialize and read vbios, but at this stage the computer nor even power up the screen or any other components. After reference clock is working, the gpu start talk with host, but vbios is already loaded by gpu, and this one is working in vesa mode. According to open source Linux documentation, there is a lot modules that require vbios itself and even more that requires by gpu core. The modules that used by driver is marked as “shared by various SW components, operations”.
Are you sure of this? After all the HP cards either don't have an EEPROM or sometimes it's completely empty.
Just verified that the Apple EFI driver does read the ATOM tables and even uses the command tables, since there's no EFI display when defining all command tables to be non-existing - a pity that it doesn't use the Object_Info table for initializing the display outputs.
But who knows, there might even be code for using that table which just has to be activated by patching the EFI driver. Do you have a tip how to find out where the EFI driver might be referring ATOM tables?
Definitely sounds like overheating problem, I have the same card and had same symptoms under heavy load (other people did not, seems not all cards behave the same).
First, make sure hardware (heatsink) installation is ok, good thermal paste on GPU and memory. Use FakeSMC kext and HWMonitor to monitor temperature diff between GPU Die and GPU Heatsink when stressing the card. Also check airflow from the ODD fan is correct (some people removed ODD bay and this changes/reduces air flow to the heatsink).
Use the 780_BR3 vbios, I think this one has the standard clock setting (not overclocked).
If you keep getting panic/reboots under stress, you can disconnect the ODD temp sensor from the motherboard, this will allow MacFansControl to spin the ODD fan up to 3800 rpm (from 2500 rpm max when temp sensor is connected), a bit more noisy but that should be enough to keep the card cool under all loads.
Hp uses oprom from system bios by whitelist injection. When even blank PCIE device is powered by host, the pci itself reporting pciid, VendorID and other data which help identify which device is connected. After that PEI program loads oprom to device and device initialize itself. But dell stores vbioses on mxm, and you cannot find oprom inside system bios. Dell laptops are easiest to upgrade but less secure. That’s why hp cards (laptop) are always blank, while wx4150 have desktop variety with Pegatron vbios flashed onboard. But if spi flash is not onboard, it’s easy to solder it in place, also need to add some 2-3 smd resistor and capacitors.
Can you share your Apple Efi research about atomtables, maybe it’s possible to make OpenCore efi patch
Thank you for your advise. I tried PRAM reset with Ethernet cable and LINUX USB Flash Stick plugged in. Situation is still the same. 😥
Code:
root@grml /lib/live/mount/persistence/sdb2/flash # ./amdvbflash -i
AMDVBFLASH version 4.71, Copyright (c) 2020 Advanced Micro Devices, Inc.
Adapter not found
1 root@grml /lib/live/mount/persistence/sdb2/flash #Trying to boot HS from HDD ... seems to boot with black screen (internal and external). Fans are quiet. Wish I had a proper 2nd system (2009 or 2010) to check whether vidcard is alive are not. I really can't believe that this card is broken as claimed to be fully tested.
Trying to get a CH341A clip programmer now ... and maybe ... if this fails too ... baking time or sending back to Asia.
Did you try Linux pci list terminal command? Lspci called or something like this
I found where the VESA mode tables are defined, separate from the StandardVESA_Timing table.
They are located in the VGA bootloader section of the vbios.
This may be an important part of the boot up sequence for HP and Dell laptops, but not in Macs. In the Mac world, the equivalent of "VESA" tables are located in the efi driver and are used for pre-boot screen initialization prior to the Mac video drivers taking over. Unfortunately, they are not neat simple tables.
They are located in the VGA bootloader section of the vbios.
This may be an important part of the boot up sequence for HP and Dell laptops, but not in Macs. In the Mac world, the equivalent of "VESA" tables are located in the efi driver and are used for pre-boot screen initialization prior to the Mac video drivers taking over. Unfortunately, they are not neat simple tables.
Last edited:
tried this now ... didn't see anything helpful ... vid card is still invisible. Checked this card again without heatsink to exclude any shorts ... still the same.
Code:
root@grml ~ # lspci
00:00.0 Host bridge: Intel Corporation 2nd Generation Core Processor Family DRAM Controller (rev 09)
00:01.0 PCI bridge: Intel Corporation Xeon E3-1200/2nd Generation Core Processor Family PCI Express Root Port (rev 09)
00:02.0 Display controller: Intel Corporation 2nd Generation Core Processor Family Integrated Graphics Controller (rev 09)
00:16.0 Communication controller: Intel Corporation 6 Series/C200 Series Chipset Family MEI Controller #1 (rev 04)
00:1a.0 USB controller: Intel Corporation 6 Series/C200 Series Chipset Family USB Universal Host Controller #5 (rev 05)
00:1a.7 USB controller: Intel Corporation 6 Series/C200 Series Chipset Family USB Enhanced Host Controller #2 (rev 05)
00:1b.0 Audio device: Intel Corporation 6 Series/C200 Series Chipset Family High Definition Audio Controller (rev 05)
00:1c.0 PCI bridge: Intel Corporation 6 Series/C200 Series Chipset Family PCI Express Root Port 1 (rev b5)
00:1c.2 PCI bridge: Intel Corporation 6 Series/C200 Series Chipset Family PCI Express Root Port 3 (rev b5)
00:1c.4 PCI bridge: Intel Corporation 6 Series/C200 Series Chipset Family PCI Express Root Port 5 (rev b5)
00:1d.0 USB controller: Intel Corporation 6 Series/C200 Series Chipset Family USB Universal Host Controller #1 (rev 05)
00:1d.7 USB controller: Intel Corporation 6 Series/C200 Series Chipset Family USB Enhanced Host Controller #1 (rev 05)
00:1f.0 ISA bridge: Intel Corporation Z68 Express Chipset LPC Controller (rev 05)
00:1f.2 SATA controller: Intel Corporation 6 Series/C200 Series Chipset Family 6 port Desktop SATA AHCI Controller (rev 05)
00:1f.3 SMBus: Intel Corporation 6 Series/C200 Series Chipset Family SMBus Controller (rev 05)
02:00.0 Ethernet controller: Broadcom Inc. and subsidiaries NetXtreme BCM57765 Gigabit Ethernet PCIe (rev 10)
03:00.0 FireWire (IEEE 1394): LSI Corporation FW643 [TrueFire] PCIe 1394b Controller (rev 08)
04:00.0 PCI bridge: Intel Corporation CV82524 Thunderbolt Controller [Light Ridge 4C 2010]
05:00.0 PCI bridge: Intel Corporation CV82524 Thunderbolt Controller [Light Ridge 4C 2010]
05:03.0 PCI bridge: Intel Corporation CV82524 Thunderbolt Controller [Light Ridge 4C 2010]
05:04.0 PCI bridge: Intel Corporation CV82524 Thunderbolt Controller [Light Ridge 4C 2010]
05:05.0 PCI bridge: Intel Corporation CV82524 Thunderbolt Controller [Light Ridge 4C 2010]
05:06.0 PCI bridge: Intel Corporation CV82524 Thunderbolt Controller [Light Ridge 4C 2010]
06:00.0 System peripheral: Intel Corporation CV82524 Thunderbolt Controller [Light Ridge 4C 2010]On the other hand, card is getting warm as I can feel temperature at heatsink after a few minutes. Have the chance to test this card in an iMac 21'' today.
Last edited:
Yes, I also tend to overheat. The radiator is installed correctly, the thermal paste with thermal pads is new. The airflow from ODD is fine, everything is in the original there. I will try to run the test and turn on the maximum ODD fan speed in macfanscontrol, I will look at the result. Thanks!
Seems like gpu is short circuit somewhere or one important smd components are broken with help of delivery services or they send it already broken . Anyway if crystal is working, then the gpu will appear in list. Dell Gpus are always comes with rom.
Code:root@grml ~ # lspci 00:00.0 Host bridge: Intel Corporation 2nd Generation Core Processor Family DRAM Controller (rev 09) 00:01.0 PCI bridge: Intel Corporation Xeon E3-1200/2nd Generation Core Processor Family PCI Express Root Port (rev 09) 00:02.0 Display controller: Intel Corporation 2nd Generation Core Processor Family Integrated Graphics Controller (rev 09) 00:16.0 Communication controller: Intel Corporation 6 Series/C200 Series Chipset Family MEI Controller #1 (rev 04) 00:1a.0 USB controller: Intel Corporation 6 Series/C200 Series Chipset Family USB Universal Host Controller #5 (rev 05) 00:1a.7 USB controller: Intel Corporation 6 Series/C200 Series Chipset Family USB Enhanced Host Controller #2 (rev 05) 00:1b.0 Audio device: Intel Corporation 6 Series/C200 Series Chipset Family High Definition Audio Controller (rev 05) 00:1c.0 PCI bridge: Intel Corporation 6 Series/C200 Series Chipset Family PCI Express Root Port 1 (rev b5) 00:1c.2 PCI bridge: Intel Corporation 6 Series/C200 Series Chipset Family PCI Express Root Port 3 (rev b5) 00:1c.4 PCI bridge: Intel Corporation 6 Series/C200 Series Chipset Family PCI Express Root Port 5 (rev b5) 00:1d.0 USB controller: Intel Corporation 6 Series/C200 Series Chipset Family USB Universal Host Controller #1 (rev 05) 00:1d.7 USB controller: Intel Corporation 6 Series/C200 Series Chipset Family USB Enhanced Host Controller #1 (rev 05) 00:1f.0 ISA bridge: Intel Corporation Z68 Express Chipset LPC Controller (rev 05) 00:1f.2 SATA controller: Intel Corporation 6 Series/C200 Series Chipset Family 6 port Desktop SATA AHCI Controller (rev 05) 00:1f.3 SMBus: Intel Corporation 6 Series/C200 Series Chipset Family SMBus Controller (rev 05) 02:00.0 Ethernet controller: Broadcom Inc. and subsidiaries NetXtreme BCM57765 Gigabit Ethernet PCIe (rev 10) 03:00.0 FireWire (IEEE 1394): LSI Corporation FW643 [TrueFire] PCIe 1394b Controller (rev 08) 04:00.0 PCI bridge: Intel Corporation CV82524 Thunderbolt Controller [Light Ridge 4C 2010] 05:00.0 PCI bridge: Intel Corporation CV82524 Thunderbolt Controller [Light Ridge 4C 2010] 05:03.0 PCI bridge: Intel Corporation CV82524 Thunderbolt Controller [Light Ridge 4C 2010] 05:04.0 PCI bridge: Intel Corporation CV82524 Thunderbolt Controller [Light Ridge 4C 2010] 05:05.0 PCI bridge: Intel Corporation CV82524 Thunderbolt Controller [Light Ridge 4C 2010] 05:06.0 PCI bridge: Intel Corporation CV82524 Thunderbolt Controller [Light Ridge 4C 2010] 06:00.0 System peripheral: Intel Corporation CV82524 Thunderbolt Controller [Light Ridge 4C 2010]
On the other hand, card is getting warm as I can feel temperature at heatsink after a few minutes. Have the chance to test this card in an iMac 21'' today.
You can try disconnect any HDD and try boot for 2-3 seconds without heatsink with dry finger on crystal. It should have temperature 40-45 on initialization after 2-3 seconds force shutdown. But if temperature goes extremely high for your finger(58-65c) you will need to make really dash shutdown. If crystal extremely hot, then the short circuit there, if ram is hot then there, if one or all Vrms are hot, then short circuit somewhere there. If you own multimeter, you can test coils with 200 ohm resistance. If you get 006-007 ohms, then something is short circuit and eating all gpu power and it cannot start. Or accidentally broke mxm female connector, or thermal paste on mxm connector
Do you remember where did you find that shorted file of firmware rom? If this is true, then it could be very useful for patching initializing stage of blank rom vbioses
This stub file is on every HP card I got. Before flashing I pull this in general from the BIOS chip and save it.
I read it directly from the EEPROM on the pictured card (twice, using CH341A), before flashing it with a custom VBIOS.
That’s great! You opened a new page of patching HP WX Gpus!!! I will make deeper research! Thanks!
How many Gpus stub files you have?