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]

Thank you for the offer.
Please allow me a little time to stabilize my production. How about we move the discussion on CPU / MEM overclock / undeclock to a new thread?

Please take your time,

You may have a comfortable sleep after you finish,
Good night.

 

[Mini_Me]

Great. Thank you very much.
So I guess we are done testing for now. I will be looking forward to see the final write up. There are so many users in multiple communities that could use a guide.

Please let me know if I could contact a few YouTubers that previously featured bios mods according to Igors Lab, and they could do a retake with these new straps and explanations.

Good morning,

I believe for easier reach, very soon I am going to make a new thread listing the vbios profiles, one for mining and one for all purposes,

I will also list the safest method to flash the bios for even moded OC bios of the dual bios card and for BE versions as well,

The previous documents that you provided before showing the results were brilliant, I hope we can use them to prepare a one document showing the best results achieved for both Micron and Samsung memories and attached to the thread and link the credit to you if you do not mind and it is fine by you,

If you have any suggestions regarding adding or improving the thread then, please feel free to share with me,

By the way, Is your system now stabilized with the desired results?

 

[AlleyCat]

Please, no need to thanks me, actually all the credits should go to you as you did and verified everything,

And you can share with whomever you like, let us spread the knowledge and let everyone get benefits from it.

Good morning,

I believe for easier reach, very soon I am going to make a new thread listing the vbios profiles, one for mining and one for all purposes,

I will also list the safest method to flash the bios for even moded OC bios of the dual bios card and for BE versions as well,

The previous documents that you provided before showing the results were brilliant, I hope we can use them to prepare a one document showing the best results achieved for both Micron and Samsung memories and attached to the thread and link the credit to you if you do not mind and it is fine by you,

If you have any suggestions regarding adding or improving the thread then, please feel free to share with me,

By the way, Is your system now stabilized with the desired results?

Good morning.
You can certainly use any information I provided, including my alias name, reports, progress, steps, and pictures in your reports. I like the way you are going to structure the report. I am looking forward to seeing the final results.

About my system, my production rig with "mixed" cards rebooted a few times last night. The Gigabyte cards are stable but running "hot" (around 117-119W). The Sapphire with Micron memory looks like the one to crash. I think the next step is to tune the power for low heat W and stability. Do you think we can test for the best power tuning? I have one Gigabyte with Micron isolated in a test rig. We can use that card for testing the lowest stable power. I am missing this power tuning process and I could not find information about it. The reports I found are about increasing the hash speed. Once we document the power tuning process we can apply it on all the cards, and we can have another valuable description for tuning a rig. Would that be OK?

 

[AlleyCat]

Good morning.
You can certainly use any information I provided, including my alias name, reports, progress, steps, and pictures in your reports. I like the way you are going to structure the report. I am looking forward to seeing the final results.

About my system, my production rig with "mixed" cards rebooted a few times last night. The Gigabyte cards are stable but running "hot" (around 117-119W). The Sapphire with Micron memory looks like the one to crash. I think the next step is to tune the power for low heat W and stability. Do you think we can test for the best power tuning? I have one Gigabyte with Micron isolated in a test rig. We can use that card for testing the lowest stable power. I am missing this power tuning process and I could not find information about it. The reports I found are about increasing the hash speed. Once we document the power tuning process we can apply it on all the cards, and we can have another valuable description for tuning a rig. Would that be OK?

One more think, I think it is also important to document how to deal with crashes. When a report shows a card has 0 MH/s, what is the best process to find what caused the crash? Is it the high memory speed? Is it the low core power? If it the low memory controller power? etc.

 

[Mini_Me]

Good morning.
You can certainly use any information I provided, including my alias name, reports, progress, steps, and pictures in your reports. I like the way you are going to structure the report. I am looking forward to seeing the final results.

About my system, my production rig with "mixed" cards rebooted a few times last night. The Gigabyte cards are stable but running "hot" (around 117-119W). The Sapphire with Micron memory looks like the one to crash. I think the next step is to tune the power for low heat W and stability. Do you think we can test for the best power tuning? I have one Gigabyte with Micron isolated in a test rig. We can use that card for testing the lowest stable power. I am missing this power tuning process and I could not find information about it. The reports I found are about increasing the hash speed. Once we document the power tuning process we can apply it on all the cards, and we can have another valuable description for tuning a rig. Would that be OK?

Good morning.
You can certainly use any information I provided, including my alias name, reports, progress, steps, and pictures in your reports. I like the way you are going to structure the report. I am looking forward to seeing the final results.

About my system, my production rig with "mixed" cards rebooted a few times last night. The Gigabyte cards are stable but running "hot" (around 117-119W). The Sapphire with Micron memory looks like the one to crash. I think the next step is to tune the power for low heat W and stability. Do you think we can test for the best power tuning? I have one Gigabyte with Micron isolated in a test rig. We can use that card for testing the lowest stable power. I am missing this power tuning process and I could not find information about it. The reports I found are about increasing the hash speed. Once we document the power tuning process we can apply it on all the cards, and we can have another valuable description for tuning a rig. Would that be OK?

Great,
Regarding tuning the power, I am glad to help, please could you provid the current used voltages and the current temperatures for the cards,

Regarding the crashes, we are going to make another post in the same thread about how to tune the mining software and how to deal with the errors and crashes related to it and the administrator of this post is going to be you as I really do not like to be alone in anything and I really would like to be a partner with you if it is totally fine from your side.

 

[AlleyCat]

Great,
Regarding tuning the power, I am glad to help, please could you provid the current used voltages and the current temperatures for the cards,

Regarding the crashes, we are going to make another post in the same thread about how to tune the mining software and how to deal with the errors and crashes related to it and the administrator of this post is going to be you as I really do not like to be alone in anything and I really would like to be a partner with you if it is totally fine from your side.

It's a pleasure working with you. The new post is at https://www.igorslab.de/community/threads/tuning-power-and-stability-for-5700xt-gpus.3366/

 

[Mini_Me]

Good morning.
You can certainly use any information I provided, including my alias name, reports, progress, steps, and pictures in your reports. I like the way you are going to structure the report. I am looking forward to seeing the final results.

About my system, my production rig with "mixed" cards rebooted a few times last night. The Gigabyte cards are stable but running "hot" (around 117-119W). The Sapphire with Micron memory looks like the one to crash. I think the next step is to tune the power for low heat W and stability. Do you think we can test for the best power tuning? I have one Gigabyte with Micron isolated in a test rig. We can use that card for testing the lowest stable power. I am missing this power tuning process and I could not find information about it. The reports I found are about increasing the hash speed. Once we document the power tuning process we can apply it on all the cards, and we can have another valuable description for tuning a rig. Would that be OK?

We will go with the simplest solution for dealing with cards running hot which is increasing the fan speed in the power table and make a new one with it as below:

For Gigabyte card:
Under Fan tab, set:
Fan Acoustic Limit 2400 RPM
Fan throttling 2400 RPM

For both Sapphire card with Micron and Samsung:
Under Fan tab, set:
Fan Acoustic Limit 1550 RPM
Fan throttling 3200 RPM

 

[Mini_Me]

It's a pleasure working with you. The new post is at https://www.igorslab.de/community/threads/tuning-power-and-stability-for-5700xt-gpus.3366/

My apologies for the misunderstanding, what I meant regarding the new post about the software tuning that its administrator is going to be you is after I make the new thread,

for turning the power for the current rig, we can continue doing it here and there was no need to make a new thread for it.

 

[Mini_Me]

It's a pleasure working with you. The new post is at https://www.igorslab.de/community/threads/tuning-power-and-stability-for-5700xt-gpus.3366/

Is it possible to change the thread title and make it for tuning the cpu and RAM?

 

[Mini_Me]

It's a pleasure working with you. The new post is at https://www.igorslab.de/community/threads/tuning-power-and-stability-for-5700xt-gpus.3366/

After I saw the pictures, too much low voltages for the long run will yield to crashes and invalid/rejected share, please increase the VDD and VDDCI to 800 mv and MVDD to 1365 mv, after that check for the stability and whether the invalid/rejected share is gone or not.

 

[AlleyCat]

Is it possible to change the thread title and make it for tuning the cpu and RAM?

Is this OK:

Tuning Power and Stability for 5700XT CPU and RAM​

 

[Mini_Me]

Is this OK:

Tuning Power and Stability for 5700XT CPU and RAM​

I mean by cpu is like AMD Ryzen cpu or Intel cpu, the system cpu.
By RAM I mean DDR4 SDRAM, the system memories modules.

For the RX 5700 XT, we are already discussing it here.

 

[AlleyCat]

I mean by cpu is like AMD Ryzen cpu or Intel cpu, the system cpu.
By RAM I mean DDR4 SDRAM, the system memories modules.

For the RX 5700 XT, we are already discussing it here.I

OK, let me move the questions about the GPU tuning to this thread.

 

[AlleyCat]

Good morning,
I want to increase the efficiency of the cards mostly by lowering the power they operate. I found that if I set the minimum core power VDD to 750mv, and the I/O bus voltage VDDCI to 750mv, some cards become unstable and the system will eventually crash.

Can you please suggest ways to find:

1) How to find the lowest power for a card?
2) How to find the lowest STABLE power for a card?
3) If a card crashes, how to find what parameter is responsible for the crash?

The first picture shows all the cards' status, and the second picture is a worksheet for testing different configurations.

I am looking forward to your reply.

Thanks,
AlleyCat

 

[AlleyCat]

We will go with the simplest solution for dealing with cards running hot which is increasing the fan speed in the power table and make a new one with it as below:

For Gigabyte card:
Under Fan tab, set:
Fan Acoustic Limit 2400 RPM
Fan throttling 2400 RPM

For both Sapphire card with Micron and Samsung:
Under Fan tab, set:
Fan Acoustic Limit 1550 RPM
Fan throttling 3200 RPM

Please let me change the values in the Power Table and test. I will focus on the Gigabyte card that is in my test rig, and because the Gigabyte cards tend to run "hot".

 

[AlleyCat]

After I saw the pictures, too much low voltages for the long run will yield to crashes and invalid/rejected share, please increase the VDD and VDDCI to 800 mv and check for the stability and whether the invalid/rejected share is gone or not.

Please find is the results of increasing the VDD and VDDCI to 800.
The increase started creating Invalid shares on the Sapphire with Samsung card.

 

[Mini_Me]

Please find is the results of increasing the VDD and VDDCI to 800.
The increase started creating Invalid shares on the Sapphire with Samsung card.

Alright let us increase the MVDD to 1365 and set VDDCI to 870 mv for both the Samsung cards.

 

[AlleyCat]

I made a change to the Card with Samsung memory only.

results: 110W
Hash: 56.64

Rig

Production

GPU1

Sapphire

Memory

Samsung

bios rev.

Modified Straps by Apple Inc

RIG

GPU

MH/s

CPU Temp.

Memory Temp

Power (W)

Core Clock (MHz)

VDD (mv)

VDDCI (mv)

MVDD (mv)

MEM (MHz)

Invalid/Crash

Ratio

Production

Sapphire

Samsung

GPU1

56.63​

42​

70​

104​

1420​

760​

760​

1350​

905​

A-710 / I-1

0.545

Production

Sapphire

Samsung

GPU1

56.64​

42​

70​

110​

1420​

800​

800​

1350​

905​

A-1031 / I-3

0.515

Production

Sapphire

Samsung

GPU1

56.34​

43​

70​

110​

109​

800​

870​

1365​

905​

0.512

 

[Mini_Me]

I made a change to the Card with Samsung memory only.

results: 110W
Hash: 56.64

Rig

Production

GPU1

Sapphire

Memory

Samsung

bios rev.

Modified Straps by Apple Inc

RIG

GPU

MH/s

CPU Temp.

Memory Temp

Power (W)

Core Clock (MHz)

VDD (mv)

VDDCI (mv)

MVDD (mv)

MEM (MHz)

Invalid/Crash

Ratio

Production

Sapphire

Samsung

GPU1

56.63​

42​

70​

104​

1420​

760​

760​

1350​

905​

A-710 / I-1

0.545

Production

Sapphire

Samsung

GPU1

56.64​

42​

70​

110​

1420​

800​

800​

1350​

905​

A-1031 / I-3

0.515

Production

Sapphire

Samsung

GPU1

56.34​

43​

70​

110​

109​

800​

870​

1365​

905​

0.512

Very well it is stable at 870 MHz, then let us set the VDDCI to 850 mv and check.

 

[AlleyCat]

Very well it is stable at 870 MHz, then let us set the VDDCI to 850 mv and check.

	GPU	MH/s	CPU Temp.	Memory Temp	Power (W)	Core Clock (MHz)	VDD (mv)	VDDCI (mv)	MVDD (mv)	MEM (MHz)	Invalid/Crash
Samsung	GPU1	56.63​	42​	70​	104​	1420​	760​	760​	1350​	905​	A-710 / I-1
Samsung	GPU1	56.64​	42​	70​	110​	1420​	800​	800​	1350​	905​	A-1031 / I-3
Samsung	GPU1	56.34​	43​	70​	110​	1420​	800​	870​	1365​	905​
Samsung	GPU1	56.63​	43​	70​	112​	1420​	800​	850​	1365​	905​