Question GIGABYTE 5700 XT Bios mod fails

[AlleyCat]

Hi,
I am failing to update Gigabyte 5700 XT. The same procedure updates the BIOS on MSI cards with no problem. I follow the instructions on Igors Lab.

The sign of having trouble with the card bios flash shows in GPU-Z. After flash using amdvbflash the values of GPU and memory frequencies is empty. With stock bios there are Mhz frequencies.

Is there any known problems with flashing Gigabyte cards?
Any suggestions on what other forums I may ask for assistance?
Is it possible that the OEM bios is signed, and any modified bios will be rejected?
If the bios is signed, any tools to resign, or would I need to buy a card from a different vendor?

Thanks,

Alley Cat

 

[Mini_Me]

Greetings all! I found this discussion from HiveOS forum (AlleyCat comment on one discussion). I can confirm that MSI gaming X is excellent choice for efficient mining. It is very sturdy as it has excellent heatsink and runs relatively cool with only two fans. I havent tried Apple straps (I'm using 1550 timings which are common knowledge and tRef = 5990 for 1550 and above. I havent changed anything with MPT. I also have to say, I recommend 5700 over XT any day over efficient mining as XT seems to be just physical 5700 with overclocking. I did flash MSI 5700 XT Gaming X bios for my card and gained zero boost on mining but it used 10W more power for no apparent reason.

Im getting these results with my MSI RX 5700 Gaming X:

Anhang anzeigen 6805

Greetings and good day,

Excellent work you have achieved there, Is the system stable on the long run? How long the system was running with these settings? And was there any system crash occurred before?

In case of system failure on the long run, you can use the parameters in the post below,

Question - GIGABYTE 5700 XT Bios mod fails

Hi, I am failing to update Gigabyte 5700 XT. The same procedure updates the BIOS on MSI cards with no problem. I follow the instructions on Igors Lab. The sign of having trouble with the card bios flash shows in GPU-Z. After flash using amdvbflash the values of GPU and memory frequencies is...

www.igorslab.de

Regarding flashing XT bios to non XT card there is no benefits because the non XT card compute units is 36 which is less than the XT one with 40 compute unites.

 

[Mini_Me]

Amazing numbers.

It will be nice if it was possible to flash non XT bios to the XT one however it is not possible.

 

[Calathia]

It will be nice if it was possible to flash non XT bios to the XT one however it is not possible.

Thank you for this info. My knowledge about differences of XT vs non XT has been incorrect.

Also forgot to say on my first post: Thank you all who have contributed for this research of yours in this thread. Its amazing value for efficient mining with 5700 / XT.

I really hope RX 6000 series will have same bios modding potential as 5000 series for new standard of efficiency on mining!

 

[Calathia]

Greetings and good day,

Excellent work you have achieved there, Is the system stable on the long run? How long the system was running with these settings? And was there any system crash occurred before?

In case of system failure on the long run, you can use the parameters in the post below,

Question - GIGABYTE 5700 XT Bios mod fails

Hi, I am failing to update Gigabyte 5700 XT. The same procedure updates the BIOS on MSI cards with no problem. I follow the instructions on Igors Lab. The sign of having trouble with the card bios flash shows in GPU-Z. After flash using amdvbflash the values of GPU and memory frequencies is...

www.igorslab.de

Regarding flashing XT bios to non XT card there is no benefits because the non XT card compute units is 36 which is less than the XT one with 40 compute unites.

It is relatively stable; it reboots itself roughly once per day (0 Mhash in some card -> rebooting). This is apparently because some cards cant handle VDD = 760 so I have changed it to 770 on these cards. Will post updates on stability after few days.

 

[Mini_Me]

Thank you for this info. My knowledge about differences of XT vs non XT has been incorrect.

Also forgot to say on my first post: Thank you all who have contributed for this research of yours in this thread. Its amazing value for efficient mining with 5700 / XT.

I really hope RX 6000 series will have same bios modding potential as 5000 series for new standard of efficiency on mining!

Your presence and kind words are really highly appreciated,

And your knowledge about the RX 5700 XT and non Xt is actually correct when you mentioned the XT and non Xt are similar however you only missed the differences between them which are the extra 4 compute units and better binning of the XT version,

The RX 6000 series will have the same modding potential of the RX 5000 series however most likely with double the results of hash rate.

 

[Mini_Me]

It is relatively stable; it reboots itself roughly once per day (0 Mhash in some card -> rebooting). This is apparently because some cards cant handle VDD = 760 so I have changed it to 770 on these cards. Will post updates on stability after few days.

Please feel free to update whenever it is possible.

 

[Mini_Me]

Amazing numbers.
I confirm that the 5700XT with your settings is reporting 111W in HiveOS.

My preferred settings on the Gigabyte 5700 XT Gaming is 58.67MH/s 115W ratio 0.51.

Please check the kindly attached images below,

 

[Mini_Me]

Greetings all! I found this discussion from HiveOS forum (AlleyCat comment on one discussion). I can confirm that MSI gaming X is excellent choice for efficient mining. It is very sturdy as it has excellent heatsink and runs relatively cool with only two fans. I havent tried Apple straps (I'm using 1550 timings which are common knowledge and tRef = 5990 for 1550 and above. I havent changed anything with MPT. I also have to say, I recommend 5700 over XT any day over efficient mining as XT seems to be just physical 5700 with overclocking. I did flash MSI 5700 XT Gaming X bios for my card and gained zero boost on mining but it used 10W more power for no apparent reason.

Im getting these results with my MSI RX 5700 Gaming X. System efficiency 0.55 as reported by HiveOS (not true reading from the wall)

Anhang anzeigen 6805

Please check the kindly attached images below,

 

[AlleyCat]

I am having troubles with the MSI/Micron card. It keeps crashing at low power. It is the only card that is not stable.
At the moment I am running the card at core 1410/800mv Mem: 930MHz 790/1350. The hash rate is 58.33 and the power is 124W. Any attempt to reduce the power by reducing VDD or VDDCI result with instability, and card crash.

Any suggestions?

Thanks,

 

[AlleyCat]

Please check the kindly attached images below,

Wow, amazing.

 

[AlleyCat]

Please check the kindly attached images below,

Can you post your modified bios for the Sapphire/Micron? I want to compare the delta between your bios and what I got to be "final"

AlleyCat

 

[AlleyCat]

I am having troubles with the MSI/Micron card. It keeps crashing at low power. It is the only card that is not stable.
At the moment I am running the card at core 1410/800mv Mem: 930MHz 790/1350. The hash rate is 58.33 and the power is 124W. Any attempt to reduce the power by reducing VDD or VDDCI result with instability, and card crash.

Any suggestions?

Thanks,

AutoFan must be on for the MSI card to work. Otherwise, the card crashes.

 

[AlleyCat]

AutoFan must be on for the MSI card to work. Otherwise, the card crashes.

The problem of MSI card crashing continues. It looked like applying an AutoFan resolved the issue. But after a restart the card crashing.

 

[AlleyCat]

The problem of MSI card crashing continues. It looked like applying an AutoFan resolved the issue. But after a restart the card crashing.

I see you are using an older version of HiveOS beta. Rev. 06-144. I am using the latest beta.
Are there any differences between the kernel or the drivers that could generate different performance results?

 

[Mini_Me]

I see you are using an older version of HiveOS beta. Rev. 06-144. I am using the latest beta.
Are there any differences between the kernel or the drivers that could generate different performance results?

Yes, There is certainly a difference, however the most important things are the kernel and adrenaline driver,

I believe the HiveOS kernel of the latest version is still 5.0.21,correct?

Please, could you inform me which Adrenaline driver version you are using, and what windows 10 version you are using as well?

I believe the issues we got here is most likely software optimization,

Regarding the MSI card, flash the original bios and only modify the vram straps, once using the Apple straps, and once using 1550 MHz straps with tREF 9998, after that check which one is more stable.

 

[Mini_Me]

Can you post your modified bios for the Sapphire/Micron? I want to compare the delta between your bios and what I got to be "final"

AlleyCat

The photos I provided before do not belong to me however to a Russian guy who have a rig of Sapphire RX 5700 XT Nitro Plus, This card proofed that it is one of the best for mining in addition to the 50th Anniversary one,

As you saw in the photos, there is a limit for the voltages optimization after that no significant gain and only instability.

By the way, please could you inform your system specification regarding the motherboard, cpu and ram?

 

[AlleyCat]

Yes, There is certainly a difference, however the most important things are the kernel and adrenaline driver,

I believe the HiveOS kernel of the latest version is still 5.0.21,correct?

Please, could you inform me which Adrenaline driver version you are using, and what windows 10 version you are using as well?

I believe the issues we got here is most likely software optimization,

Regarding the MSI card, flash the original bios and only modify the vram straps, once using the Apple straps, and once using 1550 MHz straps with tREF 9998, after that check which one is more stable.

The kernel in the latest HiveOS looks like 5.6. This is why I’m looking to get an older image for testing with an older kernel.

I will test the MSI later today.
It looks like the Nitro+ has tight tolerance components that allow them to run at higher efficiency. I thought the picture is yours and that you were able to run the Pulse at these high efficiency. Perhaps we are reaching the best efficiency with a cheaper card.

 

[Mini_Me]

The kernel in the latest HiveOS looks like 5.6. This is why I’m looking to get an older image for testing with an older kernel.

I will test the MSI later today.
It looks like the Nitro+ has tight tolerance components that allow them to run at higher efficiency. I thought the picture is yours and that you were able to run the Pulse at these high efficiency. Perhaps we are reaching the best efficiency with a cheaper card.

I believe the Nitro plus have the PCB and components of the 50th Anniversary version,

I was always suspecting the software, let us tune yours,

Is it possible to know your system specification regarding the motherboard, cpu, ram and windows 10 version?

 

[pigyang]

I read this
your good work
Sapphire RX 5700 XT Nitro Plus is very good work card
I have some Samsung reference card 5700xt
Not work well
Currently learning

 

[AlleyCat]

Yes, There is certainly a difference, however the most important things are the kernel and adrenaline driver,

I believe the HiveOS kernel of the latest version is still 5.0.21,correct?

Please, could you inform me which Adrenaline driver version you are using, and what windows 10 version you are using as well?

I believe the issues we got here is most likely software optimization,

Regarding the MSI card, flash the original bios and only modify the vram straps, once using the Apple straps, and once using 1550 MHz straps with tREF 9998, after that check which one is more stable.

I am going to modify the straps on the MSI to see if th e memory timing is the cause for crashes.