vega 56 sapphire pulse bios 1 write fail

[H3llSp4wN]

can we flash vega56 bios under linux ?

 

[eidairaman1]

can we flash vega56 bios under linux ?

No tool for Linux.

Get an SPI flasher read this for links to obtain one

https://www.techpowerup.com/forums/threads/bios-980-gtx-ti-flash.249644/page-2#post-3952037

 

[Szb84]

I have the same issue. It likely happens if you flash a V64 bios with different product/board name.

This is the default Vega56 Pulse product name: Vega10 A1 XL D06501 8GB 852e/800m 0.9V
This is how the Nitro64 and (almost) every other 56/64 bios looks: Vega10 A1 XT D05001 32Mx128 852e/945m 0.95V

The XFX Vega 64 Double Edition has a similar product name according to the bios (Vega10 A1 XT D06502 8GB 852e/945m 0.95V) and I know someone who managed to flash back the Pulse bios after flashing this one.

Still not sure why it's an issue, but this might be the cause and explain why one BIOS allows flashback and others aren't.

 

[H3llSp4wN]

Im gonna try and post Feedback!

Thanks

 

[Szb84]

Not sure it will help if the damage is already done. I for one can't flash anything on that chip after the Nitro64 rom. So either the "wrong" biosfile is causing it or there is a limit how many times you can flash it/another safety mechanism.

 

[MrGenius]

I have the same issue. It likely happens if you flash a V64 bios with different product/board name.

This is the default Vega56 Pulse product name: Vega10 A1 XL D06501 8GB 852e/800m 0.9V
This is how the Nitro64 and (almost) every other 56/64 bios looks: Vega10 A1 XT D05001 32Mx128 852e/945m 0.95V

The XFX Vega 64 Double Edition has a similar product name according to the bios (Vega10 A1 XT D06502 8GB 852e/945m 0.95V) and I know someone who managed to flash back the Pulse bios after flashing this one.

Still not sure why it's an issue, but this might be the cause and explain why one BIOS allows flashback and others aren't.

No. It's not that. That's not the product/board name. It's the config file string. And it doesn't mean anything. It can be whatever. Doesn't matter. On BIOSes that are able to be edited I change that just for fun. Doesn't make any difference. It's just info provided to identify one BIOS from another. It's doesn't actually do anything.

All of this can be changed to whatever you want, without affecting anything.

Not sure it will help if the damage is already done. I for one can't flash anything on that chip after the Nitro64 rom. So either the "wrong" biosfile is causing it or there is a limit how many times you can flash it/another safety mechanism.

It's not a thing in any practical sense. These chips are rated for >100,000 erase/program cycles per sector(industry standard spec). So they should pretty much last forever(as in survive more erase/program cycles than you'll ever manage to throw at them).

There's your limit on how many times you can flash it. On the "safety mechanism"...it's exactly the opposite of that...if anything. Not being able to flash the BIOS is not a "safe" thing if that's what needs to be done. I don't think it has anything to do with the "wrong" BIOS file either. See forthcoming explanation as to why I don't think that.

Anyhow...I don't know exactly what's going on with it. But there's one thing that's different between one Vega and another, as far as the BIOS, that seems like a likely candidate to me. It's the size of the EEPROM chip. The EEPROM chip is twice the size(2Mb vs 4Mb) on some of these Vegas that are refusing to be flashed again after a different BIOS has been flashed. Which leads me to believe it's some kind of programming error. More specifically ATIFlash is trying to flash/program a 256KB BIOS file as a 512KB BIOS file. The screenshots in this thread demonstrate what I'm talking about. 80000h bytes is 512KB when converted from hexadecimal. The actual size of the BIOS file is 40000h bytes, or 256KB when converted from hexadecimal. I presume ATIFlash is assuming the file size needs to be 512KB based on the 4Mb size of the chip, and/or because there's some other parameter it uses to determine that. So it's doing what it does, in instances when it thinks it needs to, and adding padding(filler bytes) to the BIOS file to equal 512KB. Which is throwing a monkey wrench into the works. That's my take on it anyway.

Bottom line: The problem is most likely ATIFlash. Which is a problem that is most likely solvable(at least in theory). Just don't use ATIFlash. I'm not sure how much difficulty would be involved(because I haven't tried it yet with this type of dual BIOS configuration), but it should be possible to use a hardware flasher instead. Where I think an issue might arise with that method is needing to have the card powered on to be able to access the EEPROM chip. I might be overthinking that though.

 

[Szb84]

I know you can change the data, but it can also mean that BIOS-es like that one will be causing the problem (not specifically because of that data). Because the XFX Bios seems to allow flashing back.

I also had backups from two sources: GPU-Z: 256 KB files and atiflash: 512 KB files. And even the 512 KB file will give me the same error.

 

[Wachsmann]

I know you can change the data, but it can also mean that BIOS-es like that one will be causing the problem (not specifically because of that data). Because the XFX Bios seems to allow flashing back.

I also had backups from two sources: GPU-Z: 256 KB files and atiflash: 512 KB files. And even the 512 KB file will give me the same error.

This. I tried both the 256kb and 512kb also, and same thing. If you check both files, the 512kb is just the 256kb with the rest filled with zeros.

Also tried the XFX bios, no go. No idea what that "cursed" 64 bios does, but it just won't leave the chip. Debating whether to bother getting a cheap hardware SPI flasher or just getting rid of the card completely.

 

[MrGenius]

Because the XFX Bios seems to allow flashing back.

Anecdotal evidence. Proven wrong in some cases(see post above).

I also had backups from two sources: GPU-Z: 256 KB files and atiflash: 512 KB files. And even the 512 KB file will give me the same error.

The 256KB files become the 512KB files when ATIFlash programs or saves them. Because...

If you check both files, the 512kb is just the 256kb with the rest filled with zeros.

Because that's what ATIFlash thinks it needs to do. Which I'm pretty sure has something to do with the size of the EEPROM chip. It doesn't matter what the size of the BIOS file is, if it's smaller than what ATIFlash thinks it needs to be, it will make it that size.

 

[Szb84]

Anecdotal evidence. Proven wrong in some cases(see post above).

As I said it won't work if someone already used another V64 bios that prevents flashing back. But it could be used to turn V56 Pulse to V64 (with higher HBM voltage) AND keep the ability to flash back - other bioses clearly won't allow it.

 

[jeffarese]

Sorry to bump old thread, but has anybody had luck on this?

I'm on the same situation, I can't flash the stock bios back.

 

[eidairaman1]

Sorry to bump old thread, but has anybody had luck on this?

I'm on the same situation, I can't flash the stock bios back.

Buy a spi flasher or rma the card.

 

[jeffarese]

is there any easier way? I don't think the vendor is going to allowme to email because of this

 

[eidairaman1]

is there any easier way? I don't think the vendor is going to allowme to email because of this

Email them