Learned a little since last visit. I flashed the POST vBIOS and erased the other; it boots. It seems the MI25 vBIOS directs the switching.
Also tried with the original MI25 vBIOS on other chip, booted as a WX9100.
EDIT: Or, the MI25 2nd vBIOS doesn’t setup memory timings so boot just haults. I don’t have the POST vBIOS unfortunately.
Well, I got the system to not hang after flashing an unsigned vBIOS.
Hmm, code 43.
Can’t help thinking I’m getting closer.
Going to see if I can find a vBIOS chunk that does display out, not getting that right now. After I try the unmodded “Apple” vBIOS.
So does the latest driver work fine other than fan control? Or do I need a special one for bog standard wx9100 bios
No special driver; I’m just trying to unlock more than WX9100 performance potential. I think my display issues were multiple, an adapter being shit, and temps over 60c… maybe.
Not sure about every possible reason, but finicky bullshit seems to not be an issue anymore. Almost worried about doing a fresh install and loosing some lingering config I’m unaware of. I’m not even using proper thermal paste.
Maybe I should cancel my order on 2 of the mi25… Holy shit
Still more compute and memory per$ with the instinct though
1 Like
Scratch that-
LOOKIE WHAT I HAVE
now I just need to basically learn how to use this clip flashers as I have no idea what I’m doing
3 Likes
Gonna try and force flash both chips with software
I hope you save, and share with me, the original vBIOS files. Your’s might be different and I mistakenly discarded one. Can probably use just about any serial programmer, or you’ll have better luck than I did using software.
1 Like
3 Likes
Sometimes seems fan current ratings are pesamistic; but I imagine you compared because you noticed something. You shouldn’t need much at idle to keep it plenty cool, unlike the Tesla with it many VRAM chips not thermally connected to the main heatsink.
I wonder if boot camp and an iGPU would allow the modded Apple vBIOS to fully function.
I mean, it’d be like the Tesla again with it’s own display output not working again.
okay so the MI25.rom is from AMDVB
and the Vega10 is from GPUZ
so idk why they’re different
https://drive.google.com/drive/folders/1dULSJPVmJRnULZrv5oAupAP4urGQKiOp?usp=sharing
VBFlash grabs some of the info from what the motherboard/UEFI reports, some is read directly. Only explanation for why I was in a situation where it told me “sub system ID mismatch” for flashing what I already flashed with the USB programmer, and “this ROM is the same” (or whatever the exact syntax was) for flashing the original MI25 ROM.
same card
Microsoft Windows [Version 10.0.22000.1098]
(c) Microsoft Corporation. All rights reserved.
C:\Users\Defualt\Downloads\atiflash_293>C:\Users\Defualt\Downloads\atiflash_293\amdvbflash.exe -i
adapter bn dn fn dID asic flash romsize test bios p/n
======= == == == ==== =============== ============== ======= ==== ==============
0 0A 00 00 6860 Vega10 M25P80 100000 pass 113-D0513506-999
C:\Users\Defualt\Downloads\atiflash_293>C:\Users\Defualt\Downloads\atiflash_293\amdvbflash.exe -ai
Adapter 0 (BN=0A, DN=00, FN=00, PCIID=00001002, SSID=00000C35)
Asic Family : Vega10
Flash Type : M25P80 (1024 KB)
Product Name is : Vega10 D05133 64Mx128 852e/945m
Bios Config File: D0513506.999
Bios P/N is : 113-D0513506-999
Bios Version : 016.001.001.000.000000
Bios Date is : 10/19/18 10:56
Image 1 – Offset 0x0
ROM header contents:
Signature 0xAA55
PCIR offset 0x02B0
PCI Data Structure
Signature PCIR
Vendor ID 0x1002
Device ID 0x6860
PCI Revision 0x00
Image size 0xFC00
Code revision 0x1001
Indicator 0x80 (last image)
Code type 0x00 (PCAT Image)
Legacy BIOS File Name D0513506.999
Legacy BIOS Part Number 113-D0513506-999
Legacy BIOS Build Number 432859
Legacy BIOS Change List 1621661
Binary BIOS_IDTF 0x301698B0
ByteCheckSum 0
C:\Users\Defualt\Downloads\atiflash_293>
Yeah, i realized, after re-reading, what you meant.