(omg)vflash | Fully Patched nvflash from X to Ada Lovelace [v5.780]

[t3rbin]

can you be more clear ?
Or visualize clear ?

I thought this value was easily verifiable with HxD.. once you make your changes to bios this value changes, it depends on everything you changed, it can increase or decrease. using the part of the text that is not of real use, you have to restore it to its original value by adding or subtracting. a new check with HxD should provide you with proof that it is back the same as before.

 

[Shadowsnight]

Not sure if it pulls it from their servers, or it is required to exist in ROM. Given it answers with gpu/compute mode.
BAR patch usually came with a newer VBIOS , but tool is able to pull data from their server.
Its an nvidia tool after all.

I disabled my adapter and got the same result, Interesting though it looks like 0x1b06 is the ID for your 1080ti and mine was 0x1E07 for the 2080ti, I wonder if there is a whitelist that can be changed somewhere.

Sadly i'm with limited/weak gear & fully decompiling nvflash eats over 32gb RAM.
Its huge.
With mapping out progress i'm only at 2/20 , or ~30/200 commands.
Its really huge.
Soo it will need some time

I have RAM if there is anyway I can help let me know

Edit:

S:00,B:01,D:00,F:00 Is the same for both our GPU's it'd be interesting to see what it shows on a 3000 or 4000 series card

 

[MELOCODI]

This is a god's work, it will be very helpful with GPUs like mine (ADLINK Quadro T1000 GDDR6 MXM-A).

 

[Mussels]

Now i just need something dumb like being able to make my 1070Ti identify as a GTX1080 (but with its GDDR5 ram) so i can SLI it with a real 1080


And probably have it run like ass, but yknow.

 

[Shadowsnight]

With this argument "--gpumode physical_display_enabled_8GB_bar1" you do enable the resize-bar????!!!!!

Hey just wondering if you can share a screenshot of this working? Also there is a empty setgpumode.txt file made when mine finishes, did your txt have anything in it?

 

[TheQuentincc]

So I ran it with --check and it just restarted the graphics driver? Is that normal? No errors, but no success messages either.

I have a Quadro P4000 8GB (MXM 3.0b | Pascal) so I don't really know what BIOS I could try since the P5000 has 16GB of VRAM.

You can try the P5000 vbios, the P5000 have similar set of memory config as the P4000, only a strap resistor on the board is selecting which memory config it's running.
It would be nice if it works, the 1250MHz limit on the P4000 is pretty constraining, this P4000 MXM was my test subject back in 2019/2020 when I was trying to mod Pascal vbios, so far when flashing the P5000 vbios, it wouldn't work.

If only my Dell Precision 7550 wasn't such a PITA to teardown in order to flash a new vbios (if it ever fail and end up in a brick), I would already have tried to mod my Quadro RTX 3000 vbios. I think that's one of the most interesting use of modded vbios, allowing "Geforce like" control on these cards (Pascal/Turing Quadro on which everything is locked, clocks, voltage, PL and so on) because it's a pure software lock with no workaround (unlike PL and voltage)

 

[Deleted member 218758]

I thought this value was easily verifiable with HxD.. once you make your changes to bios this value changes, it depends on everything you changed, it can increase or decrease. using the part of the text that is not of real use, you have to restore it to its original value by adding or subtracting. a new check with HxD should provide you with proof that it is back the same as before.

Sorry, theory is one thing.
It doesn't function - its not "any text".
Basic small checksums were maxwell days - tracked down TDP Tweaker's location , and offset compared for consumer GPUs
It doesn't function

Maybe user error, probably am missing something
But theory wont push things further.
Please provide examples, so people who follow this thread can use that information.

I disabled my adapter and got the same result, Interesting though it looks like 0x1b06 is the ID for your 1080ti and mine was 0x1E07 for the 2080ti, I wonder if there is a whitelist that can be changed somewhere.

I think it is not really nvflash's fault
It tries to set a gpu compute mode and fails. Likely as ROM has nothing
I'll try to track that down and see what happens - but i think its just ROM doesnt knowing the existance of BAR.
BAR for example on 3000 series came with VBIOS XX.XX.42.
38 & 39's had the potential functionally but it was disabled.
94.02.28 didnt have it.

I made bios rtx 2080 Super Bigben.

Did you change cert names and identifiers or did you edit more and replaced cert.
What did you do ?

Log verify, does not want to flash INFOROM.

Parts of InfoRom can be exported with

I checked my logs and inforom is also preserved for me on stock.
Maybe it didnt detect a missmatch. Or its my modification skipping the replacement.
I know we can backup, change and push inforom sections - where it will fix CS itself.
There must also be a force override flag for infrom, like there is a flag to wake falcon

As for that,
i haven't really noticed specific changes there yet.
Mostly its holding Boardpartners card identifiers.
So far nothing special in there, to my eyes.

 

[rootuser123]

Result of flashing RTX 2080 Super Aorus BIOS on a RTX 2070 Super Aorus. Does the debug log output indicate the blocking factor is FALCON and not NVFlash?

omgvflash -6 GV-N208SAORUS-8GC_P.rom -L con


NVIDIA driver is not present


NVIDIA driver is not present
*** NVFLASH LOG FILE ***
Command-line : -6 GV-N208SAORUS-8GC_P.rom -L con
NVIDIA Firmware Update Utility (Version 5.780.0)
Copyright (C) 1993-2022, NVIDIA Corporation. Veii Extreme Edition


NVIDIA Firmware Update Utility (Version 5.780.0)
Copyright (C) 1993-2022, NVIDIA Corporation. Veii Extreme Edition


Command: flash
FindDevices:
Device:0000:00:00:00=8086:3E30:1458:5000
Device:0000:00:01:00=8086:1901:FFFF:FFFF
Device:0000:00:08:00=8086:1911:1458:5000
Device:0000:00:14:00=8086:A2AF:1458:5007
Device:0000:00:16:00=8086:A2BA:1458:1C3A
Device:0000:00:17:00=8086:2822:1458:B005
Device:0000:00:1B:00=8086:A2EB:FFFF:FFFF
Device:0000:00:1C:00=8086:A290:FFFF:FFFF
Device:0000:00:1C:01=8086:A291:FFFF:FFFF
Device:0000:00:1C:02=8086:A292:FFFF:FFFF
Device:0000:00:1C:04=8086:A294:FFFF:FFFF
Device:0000:00:1C:06=8086:A296:FFFF:FFFF
Device:0000:00:1D:00=8086:A298:FFFF:FFFF
Device:0000:00:1F:00=8086:A2C9:1458:5001
Device:0000:00:1F:02=8086:A2A1:1458:5001
Device:0000:00:1F:03=8086:A2F0:1458:A0C1
Device:0000:01:00:00=10DE:1E84:1458:3FF6 GPU
Device:0000:01:00:01=10DE:10F8:1458:3FF6
Device:0000:01:00:02=10DE:1AD8:1458:3FF6
Device:0000:01:00:03=10DE:1AD9:1458:3FF6
Device:0000:05:00:00=1969:E0B1:1458:E000
Device:0000:06:00:00=1B21:2142:1458:5007
Device:0000:07:00:00=1B21:2142:1458:5007
Filter-out on-motherboard PLX bridge device.
No PLX/PEX switch was detected, skip to filter-out non-NVIDIA PLX/PEX switch.
One to one mode
Checking for matches between display adapter(s) and image(s)...


Testing adapter: Graphics Device      (10DE,1E84,1458,3FF6) S:00,B:01,D:00,F:00


Disable NV driver before start if NV driver loaded.


Disabling driver: Device Driver Status code = 28
Failed to raise ROM and uC perf clocks.
Flashing on blank EEPROMs will be slow.
Adapter: Graphics Device      (10DE,1E84,1458,3FF6) S:00,B:01,D:00,F:00


Identify EEPROM.
Action: Identify EEPROM.
uproc 05 ucode is not found for this GPU.
uproc 02 ucode is not found for this GPU.
uproc 04 ready to be used.
Preparing secure operation after kicked start.
Actual SPI clock frequency (Hz): 18000000
Serial EEPROM identified.
EEPROM ID (9D,7014) : ISSI IS25WP080 1.65-1.95V 8192Kx1S, page


Action: Identify EEPROM.
Serial EEPROM identified.
EEPROM ID (9D,7014) : ISSI IS25WP080 1.65-1.95V 8192Kx1S, page


Action: Read EEPROM image.
Reading EEPROM bytes.
Start: 0x00000000
Count: 1048576(0x00100000)
Read complete.=====================================] 100 %


Testing Image: GV-N208SAORUS-8GC_P.rom
IFR image mode
Invalid OTA signed VBIOS firmware image.
Current      - Parsed Production VBIOS version: 90.04.76.00.D4
Replace with - Parsed Production VBIOS version: 90.04.7A.40.3D


Mismatch Type: Computed: 0x00038007
WARNING: None of the firmware image compatible PCI Device ID's
WARNING: None of the firmware image compatible PCI Device ID's
match the PCI Device ID of the adapter.
match the PCI Device ID of the adapter.
  Adapter PCI Device ID:        1E84
  Adapter PCI Device ID:        1E84
  Firmware image PCI Device ID: 1E81
  Firmware image PCI Device ID: 1E81
    Alternate:                1EAE
    Alternate:                1EAE
WARNING: Firmware image PCI Subsystem ID (1458.3FF4)
WARNING: Firmware image PCI Subsystem ID (1458.3FF4)
  does not match adapter PCI Subsystem ID (1458.3FF6).
  does not match adapter PCI Subsystem ID (1458.3FF6).
WARNING: None of the firmware image compatible Board ID's
WARNING: None of the firmware image compatible Board ID's
match the Board ID of the adapter.
match the Board ID of the adapter.
  Adapter Board ID:        016C
  Adapter Board ID:        016C
  Firmware image Board ID: 0166
  Firmware image Board ID: 0166


Downgrading/Sidegrading firwmare detected.
(This is usually only needed for very special circumstances.)


Are you sure you want to continue?
Type "YES" to confirm (all caps):




Downgrading/Sidegrading firwmare detected.
(This is usually only needed for very special circumstances.)


Are you sure you want to continue?
Type "YES" to confirm (all caps):


YES




Downgrade/Sidegrade is permitted.
Downgrade/Sidegrade is permitted.


GPU PCI Device ID override detected.
(This is usually only needed for very special circumstances.)


Are you sure you want to continue?
Type "YES" to confirm (all caps):


GPU PCI Device ID override detected.
(This is usually only needed for very special circumstances.)


Are you sure you want to continue?
Type "YES" to confirm (all caps):
YES


Overriding the GPU PCI Device ID mismatch.
Overriding the GPU PCI Device ID mismatch.


You are intending to override PCI Subsystem ID.
Are you sure you want to continue?
Press 'y' to confirm (any other key to abort):
You are intending to override PCI Subsystem ID.
Are you sure you want to continue?
Press 'y' to confirm (any other key to abort): y
Overriding the PCI Subsystem ID mismatch.
Overriding the PCI Subsystem ID mismatch.


Board ID override detected.
*** WARNING: Overriding the Board ID can be very dangerous. ***
Upgrading to an image with the wrong Board ID might cause this system
to become or unstable.
Overriding the Board ID is only needed for extreme circumstances.
A mismatched Board ID almost always means the wrong firmware image is
being used for the specific switch.


Are you sure you want to continue?
Type "YES" to confirm (all caps):


Board ID override detected.
*** WARNING: Overriding the Board ID can be very dangerous. ***
Upgrading to an image with the wrong Board ID might cause this system
to become or unstable.
Overriding the Board ID is only needed for extreme circumstances.
A mismatched Board ID almost always means the wrong firmware image is
being used for the specific switch.


Are you sure you want to continue?
Type "YES" to confirm (all caps):
YES




Overriding the Board ID mismatch.
Overriding the Board ID mismatch.
Mismatch Type: Adjusted: 0x00000000


Current      - Version:90.04.76.00.D4 ID:10DE:1E84:1458:3FF6
               GPU Board (Normal Board)
Current      - Version:90.04.76.00.D4 ID:10DE:1E84:1458:3FF6
               GPU Board (Normal Board)
Replace with - Version:90.04.7A.40.3D ID:10DE:1E81:1458:3FF4
               GPU Board (Normal Board)
Replace with - Version:90.04.7A.40.3D ID:10DE:1E81:1458:3FF4
               GPU Board (Normal Board)


Update display adapter firmware?
Press 'y' to confirm (any other key to abort):
Action: Flash image.
Action: Identify EEPROM.
Serial EEPROM identified.
EEPROM ID (9D,7014) : ISSI IS25WP080 1.65-1.95V 8192Kx1S, page


Action: Read EEPROM image.
Reading EEPROM bytes.
Start: 0x00000000
Count: 1048576(0x00100000)
Read complete.=====================================] 100 %
G001.0000.02.04
G001.0000.02.04
InfoROM Release Level: 0x00(Current) v.s.0x00(File)
   NOTES: GENERIC_DEFAULT = 0, GENERIC_POPULATED = 1
          PRODUCT_DEFAULT = 2 ,PRODUCT_POPULATED = 3
InfoROM Release Version: G001.0000.02.04(Current) v.s. G001.0000.02.04(File)
INFOROM_FLASH_TYPE: 0x05
INFOROM_FLASH_TYPE_GENERIC_DEFAULT_TO_GENERIC_DEFAULT
Refresh InfoROM Backup (IB) block from the source vbios in EEPROM.
Preserve License image...
Preserve InfoROM OBD object...
Succeed to preserve InfoROM OBD.
Preserve InfoROM OEM object...
Succeed to preserve InfoROM OEM object.
Preserve InfoROM GXI object...
GXI preservation completes.
Preserve InfoROM RPR object...
RPR preservation completes.
Preserve InfoROM PBL object...
Skip to preserve InfoROM PBL data for EEPROM does not have PBL data.
Preserve InfoROM BBO object...
Succeed to preserve InfoROM BBO object.
Leverage current InfoROM offset and remove the existed InfoROM before composing new InfoROM
InfoROM PCI base offset: Current offset = 0xF9E00
                         New offset = 0xF9E00
Final PCI expansion ROM size is 23552(0x5C00) bytes
Final InfoROM Offset = 0x000D1A00 ( Absolute Offset 0x000FA000)
RFRD Settings    | ROMImage   | Final      |
InfoRomOffset    | 0x000FA000 | 0x000FA000 |
InfoRomSize      | 0x00004000 | 0x00004000 |


InfoROM Offsets    | ROMImage | InputAlt | Final    |
NBSI OpRom Offset  | 0x0F9E00 | 0x000000 | 0x0F9E00 |
NPDS Struct Offset | 0x0F9E20 | 0x000020 | 0x0F9E20 |
NPDE Offset        | 0x0F9E40 | 0x000040 | 0x0F9E40 |
NBSI DIR Offset    | 0x0F9FE0 | 0x000FE0 | 0x0F9FE0 |
JFFS Offset        | 0x0FA000 | 0x001000 | 0x0FA000 |
Delete the current InfoROM block from the target vbios image.
Num of Objects = 7
Num of RW Objects = 2
Size of image with dummy RO blocks  = 2688
Leverage current InfoROM offset and remove the existed InfoROM before composing new InfoROM
InfoROM PCI base offset: Current offset = 0xF9E00
                         New offset = 0xF9E00
Final PCI expansion ROM size is 23552(0x5C00) bytes
Final InfoROM Offset = 0x000D1A00 ( Absolute Offset 0x000FA000)
RFRD Settings    | ROMImage   | Final      |
InfoRomOffset    | 0x000FA000 | 0x000FA000 |
InfoRomSize      | 0x00004000 | 0x00004000 |


InfoROM Offsets    | ROMImage | InputAlt | Final    |
NBSI OpRom Offset  | 0x0F9E00 | 0x000000 | 0x0F9E00 |
NPDS Struct Offset | 0x0F9E20 | 0x000020 | 0x0F9E20 |
NPDE Offset        | 0x0F9E40 | 0x000040 | 0x0F9E40 |
NBSI DIR Offset    | 0x0F9FE0 | 0x000FE0 | 0x0F9FE0 |
JFFS Offset        | 0x0FA000 | 0x001000 | 0x0FA000 |
Num of Objects = 7
Num of RW Objects = 2
Size of image with dummy RO blocks  = 2688
Num of Objects = 7
Num of RW Objects = 2
Size of image with dummy RO blocks  = 2688
Num of Objects = 7
Num of RW Objects = 2
Size of image with dummy RO blocks  = 2688
Skip proceeding with preservation table for Pascal or newer chip.
EEPROM    size: 00100000
Image     size: 000FFA00
Unused    size: 00000600
Filled    size: 00000600


BCRT: Start Certificate 2.0 verification
Send VV Command...
cmd: ROMDIR
Cert info block will be finalized during flash process.
Program page  Start: 0x00000000 Count: 0x2000(8192)
Command id: 0x30000005 Command: NV_UCODE_CMD_COMMAND_EWR failed
Command Status: NV_UCODE_CMD_STS_COMPLETE
Error Code = 0x0000005A(90): NV_UCODE_ERR_CODE_DEVID_MATCH_LIST_DEVID_MATCH_FAILED
Device ID in the rom image does not match the device ID in the chip.
Update aborted, please use a rom image that supports device ID 0x1E84.


Falcon CLOSE


EEPROM programming failed.


NOTE: Exception caught.
Nothing changed!
Nothing changed!












PROGRAMMING ERROR: EEPROM Program failed
PROGRAMMING ERROR: EEPROM Program failed
Exit Code: 4

 

[Space Lynx]

Result of flashing RTX 2080 Super Aorus BIOS on a RTX 2070 Super Aorus. Does the debug log output indicate the blocking factor is FALCON and not NVFlash?

omgvflash -6 GV-N208SAORUS-8GC_P.rom -L con


NVIDIA driver is not present


NVIDIA driver is not present
*** NVFLASH LOG FILE ***
Command-line : -6 GV-N208SAORUS-8GC_P.rom -L con
NVIDIA Firmware Update Utility (Version 5.780.0)
Copyright (C) 1993-2022, NVIDIA Corporation. Veii Extreme Edition


NVIDIA Firmware Update Utility (Version 5.780.0)
Copyright (C) 1993-2022, NVIDIA Corporation. Veii Extreme Edition


Command: flash
FindDevices:
Device:0000:00:00:00=8086:3E30:1458:5000
Device:0000:00:01:00=8086:1901:FFFF:FFFF
Device:0000:00:08:00=8086:1911:1458:5000
Device:0000:00:14:00=8086:A2AF:1458:5007
Device:0000:00:16:00=8086:A2BA:1458:1C3A
Device:0000:00:17:00=8086:2822:1458:B005
Device:0000:00:1B:00=8086:A2EB:FFFF:FFFF
Device:0000:00:1C:00=8086:A290:FFFF:FFFF
Device:0000:00:1C:01=8086:A291:FFFF:FFFF
Device:0000:00:1C:02=8086:A292:FFFF:FFFF
Device:0000:00:1C:04=8086:A294:FFFF:FFFF
Device:0000:00:1C:06=8086:A296:FFFF:FFFF
Device:0000:00:1D:00=8086:A298:FFFF:FFFF
Device:0000:00:1F:00=8086:A2C9:1458:5001
Device:0000:00:1F:02=8086:A2A1:1458:5001
Device:0000:00:1F:03=8086:A2F0:1458:A0C1
Device:0000:01:00:00=10DE:1E84:1458:3FF6 GPU
Device:0000:01:00:01=10DE:10F8:1458:3FF6
Device:0000:01:00:02=10DE:1AD8:1458:3FF6
Device:0000:01:00:03=10DE:1AD9:1458:3FF6
Device:0000:05:00:00=1969:E0B1:1458:E000
Device:0000:06:00:00=1B21:2142:1458:5007
Device:0000:07:00:00=1B21:2142:1458:5007
Filter-out on-motherboard PLX bridge device.
No PLX/PEX switch was detected, skip to filter-out non-NVIDIA PLX/PEX switch.
One to one mode
Checking for matches between display adapter(s) and image(s)...


Testing adapter: Graphics Device      (10DE,1E84,1458,3FF6) S:00,B:01,D:00,F:00


Disable NV driver before start if NV driver loaded.


Disabling driver: Device Driver Status code = 28
Failed to raise ROM and uC perf clocks.
Flashing on blank EEPROMs will be slow.
Adapter: Graphics Device      (10DE,1E84,1458,3FF6) S:00,B:01,D:00,F:00


Identify EEPROM.
Action: Identify EEPROM.
uproc 05 ucode is not found for this GPU.
uproc 02 ucode is not found for this GPU.
uproc 04 ready to be used.
Preparing secure operation after kicked start.
Actual SPI clock frequency (Hz): 18000000
Serial EEPROM identified.
EEPROM ID (9D,7014) : ISSI IS25WP080 1.65-1.95V 8192Kx1S, page


Action: Identify EEPROM.
Serial EEPROM identified.
EEPROM ID (9D,7014) : ISSI IS25WP080 1.65-1.95V 8192Kx1S, page


Action: Read EEPROM image.
Reading EEPROM bytes.
Start: 0x00000000
Count: 1048576(0x00100000)
Read complete.=====================================] 100 %


Testing Image: GV-N208SAORUS-8GC_P.rom
IFR image mode
Invalid OTA signed VBIOS firmware image.
Current      - Parsed Production VBIOS version: 90.04.76.00.D4
Replace with - Parsed Production VBIOS version: 90.04.7A.40.3D


Mismatch Type: Computed: 0x00038007
WARNING: None of the firmware image compatible PCI Device ID's
WARNING: None of the firmware image compatible PCI Device ID's
match the PCI Device ID of the adapter.
match the PCI Device ID of the adapter.
  Adapter PCI Device ID:        1E84
  Adapter PCI Device ID:        1E84
  Firmware image PCI Device ID: 1E81
  Firmware image PCI Device ID: 1E81
    Alternate:                1EAE
    Alternate:                1EAE
WARNING: Firmware image PCI Subsystem ID (1458.3FF4)
WARNING: Firmware image PCI Subsystem ID (1458.3FF4)
  does not match adapter PCI Subsystem ID (1458.3FF6).
  does not match adapter PCI Subsystem ID (1458.3FF6).
WARNING: None of the firmware image compatible Board ID's
WARNING: None of the firmware image compatible Board ID's
match the Board ID of the adapter.
match the Board ID of the adapter.
  Adapter Board ID:        016C
  Adapter Board ID:        016C
  Firmware image Board ID: 0166
  Firmware image Board ID: 0166


Downgrading/Sidegrading firwmare detected.
(This is usually only needed for very special circumstances.)


Are you sure you want to continue?
Type "YES" to confirm (all caps):




Downgrading/Sidegrading firwmare detected.
(This is usually only needed for very special circumstances.)


Are you sure you want to continue?
Type "YES" to confirm (all caps):


YES




Downgrade/Sidegrade is permitted.
Downgrade/Sidegrade is permitted.


GPU PCI Device ID override detected.
(This is usually only needed for very special circumstances.)


Are you sure you want to continue?
Type "YES" to confirm (all caps):


GPU PCI Device ID override detected.
(This is usually only needed for very special circumstances.)


Are you sure you want to continue?
Type "YES" to confirm (all caps):
YES


Overriding the GPU PCI Device ID mismatch.
Overriding the GPU PCI Device ID mismatch.


You are intending to override PCI Subsystem ID.
Are you sure you want to continue?
Press 'y' to confirm (any other key to abort):
You are intending to override PCI Subsystem ID.
Are you sure you want to continue?
Press 'y' to confirm (any other key to abort): y
Overriding the PCI Subsystem ID mismatch.
Overriding the PCI Subsystem ID mismatch.


Board ID override detected.
*** WARNING: Overriding the Board ID can be very dangerous. ***
Upgrading to an image with the wrong Board ID might cause this system
to become or unstable.
Overriding the Board ID is only needed for extreme circumstances.
A mismatched Board ID almost always means the wrong firmware image is
being used for the specific switch.


Are you sure you want to continue?
Type "YES" to confirm (all caps):


Board ID override detected.
*** WARNING: Overriding the Board ID can be very dangerous. ***
Upgrading to an image with the wrong Board ID might cause this system
to become or unstable.
Overriding the Board ID is only needed for extreme circumstances.
A mismatched Board ID almost always means the wrong firmware image is
being used for the specific switch.


Are you sure you want to continue?
Type "YES" to confirm (all caps):
YES




Overriding the Board ID mismatch.
Overriding the Board ID mismatch.
Mismatch Type: Adjusted: 0x00000000


Current      - Version:90.04.76.00.D4 ID:10DE:1E84:1458:3FF6
               GPU Board (Normal Board)
Current      - Version:90.04.76.00.D4 ID:10DE:1E84:1458:3FF6
               GPU Board (Normal Board)
Replace with - Version:90.04.7A.40.3D ID:10DE:1E81:1458:3FF4
               GPU Board (Normal Board)
Replace with - Version:90.04.7A.40.3D ID:10DE:1E81:1458:3FF4
               GPU Board (Normal Board)


Update display adapter firmware?
Press 'y' to confirm (any other key to abort):
Action: Flash image.
Action: Identify EEPROM.
Serial EEPROM identified.
EEPROM ID (9D,7014) : ISSI IS25WP080 1.65-1.95V 8192Kx1S, page


Action: Read EEPROM image.
Reading EEPROM bytes.
Start: 0x00000000
Count: 1048576(0x00100000)
Read complete.=====================================] 100 %
G001.0000.02.04
G001.0000.02.04
InfoROM Release Level: 0x00(Current) v.s.0x00(File)
   NOTES: GENERIC_DEFAULT = 0, GENERIC_POPULATED = 1
          PRODUCT_DEFAULT = 2 ,PRODUCT_POPULATED = 3
InfoROM Release Version: G001.0000.02.04(Current) v.s. G001.0000.02.04(File)
INFOROM_FLASH_TYPE: 0x05
INFOROM_FLASH_TYPE_GENERIC_DEFAULT_TO_GENERIC_DEFAULT
Refresh InfoROM Backup (IB) block from the source vbios in EEPROM.
Preserve License image...
Preserve InfoROM OBD object...
Succeed to preserve InfoROM OBD.
Preserve InfoROM OEM object...
Succeed to preserve InfoROM OEM object.
Preserve InfoROM GXI object...
GXI preservation completes.
Preserve InfoROM RPR object...
RPR preservation completes.
Preserve InfoROM PBL object...
Skip to preserve InfoROM PBL data for EEPROM does not have PBL data.
Preserve InfoROM BBO object...
Succeed to preserve InfoROM BBO object.
Leverage current InfoROM offset and remove the existed InfoROM before composing new InfoROM
InfoROM PCI base offset: Current offset = 0xF9E00
                         New offset = 0xF9E00
Final PCI expansion ROM size is 23552(0x5C00) bytes
Final InfoROM Offset = 0x000D1A00 ( Absolute Offset 0x000FA000)
RFRD Settings    | ROMImage   | Final      |
InfoRomOffset    | 0x000FA000 | 0x000FA000 |
InfoRomSize      | 0x00004000 | 0x00004000 |


InfoROM Offsets    | ROMImage | InputAlt | Final    |
NBSI OpRom Offset  | 0x0F9E00 | 0x000000 | 0x0F9E00 |
NPDS Struct Offset | 0x0F9E20 | 0x000020 | 0x0F9E20 |
NPDE Offset        | 0x0F9E40 | 0x000040 | 0x0F9E40 |
NBSI DIR Offset    | 0x0F9FE0 | 0x000FE0 | 0x0F9FE0 |
JFFS Offset        | 0x0FA000 | 0x001000 | 0x0FA000 |
Delete the current InfoROM block from the target vbios image.
Num of Objects = 7
Num of RW Objects = 2
Size of image with dummy RO blocks  = 2688
Leverage current InfoROM offset and remove the existed InfoROM before composing new InfoROM
InfoROM PCI base offset: Current offset = 0xF9E00
                         New offset = 0xF9E00
Final PCI expansion ROM size is 23552(0x5C00) bytes
Final InfoROM Offset = 0x000D1A00 ( Absolute Offset 0x000FA000)
RFRD Settings    | ROMImage   | Final      |
InfoRomOffset    | 0x000FA000 | 0x000FA000 |
InfoRomSize      | 0x00004000 | 0x00004000 |


InfoROM Offsets    | ROMImage | InputAlt | Final    |
NBSI OpRom Offset  | 0x0F9E00 | 0x000000 | 0x0F9E00 |
NPDS Struct Offset | 0x0F9E20 | 0x000020 | 0x0F9E20 |
NPDE Offset        | 0x0F9E40 | 0x000040 | 0x0F9E40 |
NBSI DIR Offset    | 0x0F9FE0 | 0x000FE0 | 0x0F9FE0 |
JFFS Offset        | 0x0FA000 | 0x001000 | 0x0FA000 |
Num of Objects = 7
Num of RW Objects = 2
Size of image with dummy RO blocks  = 2688
Num of Objects = 7
Num of RW Objects = 2
Size of image with dummy RO blocks  = 2688
Num of Objects = 7
Num of RW Objects = 2
Size of image with dummy RO blocks  = 2688
Skip proceeding with preservation table for Pascal or newer chip.
EEPROM    size: 00100000
Image     size: 000FFA00
Unused    size: 00000600
Filled    size: 00000600


BCRT: Start Certificate 2.0 verification
Send VV Command...
cmd: ROMDIR
Cert info block will be finalized during flash process.
Program page  Start: 0x00000000 Count: 0x2000(8192)
Command id: 0x30000005 Command: NV_UCODE_CMD_COMMAND_EWR failed
Command Status: NV_UCODE_CMD_STS_COMPLETE
Error Code = 0x0000005A(90): NV_UCODE_ERR_CODE_DEVID_MATCH_LIST_DEVID_MATCH_FAILED
Device ID in the rom image does not match the device ID in the chip.
Update aborted, please use a rom image that supports device ID 0x1E84.


Falcon CLOSE


EEPROM programming failed.


NOTE: Exception caught.
Nothing changed!
Nothing changed!












PROGRAMMING ERROR: EEPROM Program failed
PROGRAMMING ERROR: EEPROM Program failed
Exit Code: 4

shouldn't it fail? how are you expecting to get a free gpu upgrade, 2070 to 2080? isn't that less cuda cores physically on the chip itself? man i don't understand this thread, either you are all way smarter than me, or I am just missing something.

also, most people I know with a 4090 limit it in fps or vsync it cause it uses so much wattage/already maxes out fps in games, so those of you going to full 1.1v 4090, i just don't know the point of that. lol

 

[rootuser123]

shouldn't it fail? how are you expecting to get a free gpu upgrade, 2070 to 2080? isn't that less cuda cores physically on the chip itself? man i don't understand this thread, either you are all way smarter than me, or I am just missing something.

also, most people I know with a 4090 limit it in fps or vsync it cause it uses so much wattage/already maxes out fps in games, so those of you going to full 1.1v 4090, i just don't know the point of that. lol

Yes, the cores are lasered off, but the 2080 Super Aorus BIOS has a higher TDP limit which mitigates PL throttling.

 

[Space Lynx]

Yes, the cores are lasered off, but the 2080 Super Aorus BIOS has a higher TDP limit which mitigates PL throttling.

thanks, that does help me understand better. my 7900 xt has no such issues, it will just get 96 celsius hotspot when heavily oc'd, remain fully stable and no issues at all. but i prefer my cards to run cold, so i keep it on the default bios with a mild oc and a more reasonable 81 celsius hotspot. not seen any kind of throttling

 

[rootuser123]

thanks, that does help me understand better. my 7900 xt has no such issues, it will just get 96 celsius hotspot when heavily oc'd, remain fully stable and no issues at all. but i prefer my cards to run cold, so i keep it on the default bios with a mild oc and a more reasonable 81 celsius hotspot. not seen any kind of throttling

I have a 7900 XT ref in my other system, and it does like a good undervolt with maxxed out PL which nets me the best benchmark scores.

 

[Space Lynx]

I have a 7900 XT ref in my other system, and it does like a good undervolt with maxxed out PL which nets me the best benchmark scores.

yeah i have a 1025 unvervolt and boost to 2800-2900 on my 7900 xt, but it won't go further than that, i mean even if you removed PL or changed it from 15% to 25%, i feel like it would just become unstable cause the silicon is already being pushed pretty damn good. but i may be wrong on that.

 

[kakaesthetic]

i flashed my 3070 ti zotac trinity to bios zotac amp extreme, but the clock still 300 in stress , i returned to original bios and is okay

Does anyone know why the gpu clock doesn't go up and stay at 300?
I'm glad I managed to go back to the original bios and went back to raising the clock to 1770 normally

ddu was used, I reinstalled the driver, literally everything possible and with the mod bios the clock did not go up, it just returned to normal after putting the original bios

 

[W1zzard]

Now i just need something dumb like being able to make my 1070Ti identify as a GTX1080

You know that you can just change the driver string in the INF file to set the device name in Windows? NVCleanstall will even help you with that. It won't unlock anything of course. Some vendors of fake chinese NVIDIA cards have been doing that. The card only works with "their" driver, which has the changed name.

 

[Pichulec]

I have tried flashing my card however it fails saying eeprom not recognized. Palit Gamerock 4080. I succeeded with Nvflashk.

 

[SakuraSamurai]

You can try the P5000 vbios, the P5000 have similar set of memory config as the P4000, only a strap resistor on the board is selecting which memory config it's running.
It would be nice if it works, the 1250MHz limit on the P4000 is pretty constraining, this P4000 MXM was my test subject back in 2019/2020 when I was trying to mod Pascal vbios, so far when flashing the P5000 vbios, it wouldn't work.

If only my Dell Precision 7550 wasn't such a PITA to teardown in order to flash a new vbios (if it ever fail and end up in a brick), I would already have tried to mod my Quadro RTX 3000 vbios. I think that's one of the most interesting use of modded vbios, allowing "Geforce like" control on these cards (Pascal/Turing Quadro on which everything is locked, clocks, voltage, PL and so on) because it's a pure software lock with no workaround (unlike PL and voltage)

--protectoff seems to have worked fine. So maybe i'll try to flash it, I can always enable the iGPU if I have to and flash back.

I agree. My Precision M6600 is one thing to take apart. labor intensive, but not tedious. I really don't want to take apart this 7720, it doesn't look fun.
I max out around 68C Hotspot when pushing this card in games or compute loads. It has so much headroom and potential. Especially since GP104 overclocks well, as does the samsung memory.

 

[dudko__SVK]

hi i try flash me RTX3090 FE whit evga 3090 ftw https://www.techpowerup.com/vgabios/234962/evga-rtx3090-24576-201209 pc wont booot up stuck on q-code 97
somebody try whit difirent and work or

 

[t3rbin]

Sorry, theory is one thing.
It doesn't function - its not "any text".
Basic small checksums were maxwell days - tracked down TDP Tweaker's location , and offset compared for consumer GPUs
It doesn't function

Maybe user error, probably am missing something
But theory wont push things further.
Please provide examples, so people who follow this thread can use that information.

oops!!! I hope we haven't already run out of ideas here and someone finds a solution to this soon.

 

[Wotwow]

Unfortunately, this is not a GSP hack.
For bios edit, a simple version of nvflash is better when they see errors.
This tool still does not allow flashing the firmware completely bypassing the GSP (Falcon).

P.S.
I'm sorry author for the truth.

Result of flashing RTX 2080 Super Aorus BIOS on a RTX 2070 Super Aorus. Does the debug log output indicate the blocking factor is FALCON and not NVFlash?

It doesn't flash.
You must save your bios.
Flash.
To restart a computer
Then through gpuz to save bios.
After:
Then compare with your bios.

This is not hack GSP.
Use the normal version of nvflash if you're good at the Hex tool.

 

[K4sum1]

Is there a good place to discuss GPU vBIOS modding? When trying to find modded vBIOSes in the past I just find them scattered around. Is there somewhere I'm missing that all this talk takes place in? I don't want to dump my questions or requests here, except one.

Could I flash a TU102-300A-K3-A1 vBIOS on a TU102-300-K3-A1 with this tool?

 

[Wotwow]

Is there a good place to discuss GPU vBIOS modding? When trying to find modded vBIOSes in the past I just find them scattered around. Is there somewhere I'm missing that all this talk takes place in? I don't want to dump my questions or requests here, except one.

Could I flash a TU102-300A-K3-A1 vBIOS on a TU102-300-K3-A1 with this tool?

Any utility will do if you use the correct BIOS.

 

[K4sum1]

Well, it was a big issue with the 2080 Ti, you couldn't flash a 300A vBIOS on a 300, even though the 300A is a binned 300.

Although if this allows for vBIOS modification, maybe it would be easier to just jack up the power and voltage limits instead of trying to force a 300A vBIOS on a 300 card.

 

[R-T-B]

Downside: welcome to the same shit AMD has dealt with and second hand GPUs being fakes/modded to death

I'm all for being able to mod our cards, but seeing the things done in bulk due to eth mining has really soured me to it.

Not me. I just see it as another reason for a return if the seller can't be arsed to flash the friggin original bios, at minimum. This isn't a modding issue it's a marketplace issue.

This is the reason I joined TPU and it pleases me to see it returning. All you young kids have no idea: This site was literally built on this shit. That's not funny, that's a fact. Ask @W1zzard Long history there.

 

[Wotwow]

Not me. I just see it as another reason for a return if the seller can't be arsed to flash the friggin original bios, at minimum. This isn't a modding issue it's a marketplace issue.

This is the reason I joined TPU and it pleases me to see it returning. All you young kids have no idea: This site was literally built on this shit. That's not funny, that's a fact. Ask @W1zzard Long history there.

There was a time when I converted gtx 780 to TITAN.
Only memory error correction did not work.
Now in Turing I can unlock the increased cache.
But can't unlock double point performance.
I thought that these tools would be able to flash the bios bypassing the check, but they can't.

Parts of InfoRom can be exported with
I checked my logs and inforom is also preserved for me on stock.
Maybe it didnt detect a missmatch. Or its my modification skipping the replacement.
I know we can backup, change and push inforom sections - where it will fix CS itself.
There must also be a force override flag for infrom, like there is a flag to wake falcon

As for that,
i haven't really noticed specific changes there yet.
Mostly its holding Boardpartners card identifiers.
So far nothing special in there, to my eyes.

I do not have a InfoRom from a quadro rtx 5000.

Sorry, theory is one thing.
What did you do ?