Mining rig keeps restarting

[minefewcoins]

why not just placing a high cfm fan over the mobo , and let it blow air to the whole motherboard ?

I found the spot. It is south bridge, and the gpu is right above it, so it does'nt have proper airflow. I put the fan near it and it lowered temps from 95+ to 81-82. Will repaste it with kryonaut. That way it should go even few C lower.

[shibob]

Just look at your GPU's temp, why you let it over 90C.
Make it cool with below 65C, your rig will be stable for 24/7 operation for sure.

[Geraldo]

I put fan on manual and it solved graphic cards problems...

Glad to hear that.

But where is this "Temperature 4" sensor on MOBO? It goes over 93C

Have you tried other system information utility software? Just to make sure that what shown on HW Info is accurate. If you think what shown in HW info is accurate, you could try this advice.

why not just placing a high cfm fan over the mobo , and let it blow air to the whole motherboard ?

[minefewcoins]

It was south bridge problem which was fixed by applying thermal paste and gpu fan-stop which was solved by setting fan to start spinning on lower temps. Temps are now arround 60C for core and 70 for HBM2
It should be stable now. Thank you all for suggestions.

[minefewcoins]

I have another problem now. Three cards worked like a charm. But when I add the forth it restarts after few minutes. Tried to change different risers, tried to change gpu, tried to change riser power cable and it's slot on psu, tried to change pci-e slot on mobo and even tried to put power cable from 4th gpu on other one. Yeah, and I tried to lower power consumption. Nothing helps. So only thing that is left is some software problem I guess. Is there some bios or wattman setting that could cause this? Got new mobo btw. It is Asus prime Z270-a. Bios update didn't help too.

[smoolae]

I found some ASUS MOBO settings you could try:

Advanced mode (F7) -> "Boot" section -> set "4G Decode" to ENABLED.
Under "Boot" -> set FAST BOOT to DISABLED.
Under "Advanced" -> Onboard Devices Configuration -> set HD AUDIO CONTROLLER to DISABLED.
Under "Advanced" -> PCH Configuration -> set PCI-E Link Speed to GEN2.

Note: Some users have reported GEN1 working better with the ASUS PRIME Z270P. Try that if you have that board.

But usually restarting screams power supply not providing enough power. Does it restart when you start the miner program or just when idling?

[minefewcoins]

I found some ASUS MOBO settings you could try:

Advanced mode (F7) -> "Boot" section -> set "4G Decode" to ENABLED.
Under "Boot" -> set FAST BOOT to DISABLED.
Under "Advanced" -> Onboard Devices Configuration -> set HD AUDIO CONTROLLER to DISABLED.
Under "Advanced" -> PCH Configuration -> set PCI-E Link Speed to GEN2.

Note: Some users have reported GEN1 working better with the ASUS PRIME Z270P. Try that if you have that board.

But usually restarting screams power supply not providing enough power. Does it restart when you start the miner program or just when idling?

Thank you for the suggestions. From this settings I only dont have FAST BOOT DISABLED. Will check that settings soon.
When idle it does not reset. I left it overnight once. When I start claymore it is ok for few minutes and then just crashes. It can be 2 mins after running claymore and can be 15 minutes. It depends, but I don't know on what. In windows event manager it gives me 6008 error. I have read that it could suggest overheating. But GPU is arround 60C. Hbm was max 75C but mostly under that and VRM gets to 80C CPU nad mobo are ice cold.

[smoolae]

I found some ASUS MOBO settings you could try:

Advanced mode (F7) -> "Boot" section -> set "4G Decode" to ENABLED.
Under "Boot" -> set FAST BOOT to DISABLED.
Under "Advanced" -> Onboard Devices Configuration -> set HD AUDIO CONTROLLER to DISABLED.
Under "Advanced" -> PCH Configuration -> set PCI-E Link Speed to GEN2.

Note: Some users have reported GEN1 working better with the ASUS PRIME Z270P. Try that if you have that board.

But usually restarting screams power supply not providing enough power. Does it restart when you start the miner program or just when idling?

Thank you for the suggestions. From this settings I only dont have FAST BOOT DISABLED. Will check that settings soon.
When idle it does not reset. I left it overnight once. When I start claymore it is ok for few minutes and then just crashes. It can be 2 mins after running claymore and can be 15 minutes. It depends, but I don't know on what. In windows event manager it gives me 6008 error. I have read that it could suggest overheating. But GPU is arround 60C. Hbm was max 75C but mostly under that and VRM gets to 80C CPU nad mobo are ice cold.

Still sound grossly like a power delivery failure. Something cannot get enough power and restarts the rig in the outcome. If you have a spare PSU you could try that.

[minefewcoins]

I found some ASUS MOBO settings you could try:

Advanced mode (F7) -> "Boot" section -> set "4G Decode" to ENABLED.
Under "Boot" -> set FAST BOOT to DISABLED.
Under "Advanced" -> Onboard Devices Configuration -> set HD AUDIO CONTROLLER to DISABLED.
Under "Advanced" -> PCH Configuration -> set PCI-E Link Speed to GEN2.

Note: Some users have reported GEN1 working better with the ASUS PRIME Z270P. Try that if you have that board.

But usually restarting screams power supply not providing enough power. Does it restart when you start the miner program or just when idling?

Thank you for the suggestions. From this settings I only dont have FAST BOOT DISABLED. Will check that settings soon.
When idle it does not reset. I left it overnight once. When I start claymore it is ok for few minutes and then just crashes. It can be 2 mins after running claymore and can be 15 minutes. It depends, but I don't know on what. In windows event manager it gives me 6008 error. I have read that it could suggest overheating. But GPU is arround 60C. Hbm was max 75C but mostly under that and VRM gets to 80C CPU nad mobo are ice cold.

Still sound grossly like a power delivery failure. Something cannot get enough power and restarts the rig in the outcome. If you have a spare PSU you could try that.

The PSU was sent at shop to check it out and it was loaded at almost max load and working 12+h without restarts. I can't test it because I don't have second PSU powerful enough.

[Raja_MBZ]

Im pretty sure they shouldn't run hotter than 80C

80 degrees is a lot. I have in the settings Claymore is off at 72 degrees. It is possible to switch-off rigs OP in the same range. I never allow my GPU to work at temperatures above 60 degrees. It is in dual mode. It's cold in my country now. Therefore, my rigs are not heated above 50 degrees. I think this is the best regime.

My rigs have been running at 88 degrees for 2 years now, yet you guys are thinking that 80+ degrees is a lot.

Around 85 degrees is just fine for GPU's in my opinion.

[Geraldo]

The PSU was sent at shop to check it out and it was loaded at almost max load and working 12+h without restarts. I can't test it because I don't have second PSU powerful enough.

it's a right decision, on a GPUs Rig troubleshooting, the physical check is important, check devices position, temperature, etc.

My rigs have been running at 88 degrees for 2 years now, yet you guys are thinking that 80+ degrees is a lot.

Around 85 degrees is just fine for GPU's in my opinion.

That's nothing with mine. I run two days my Gigabyte Windforce 290X with the temperature between 91-92, I stop it cause my PSU was burning. Now, that card run normally, under 85 Lol. Yeah, it's normal if I compare it with the temperature before.

[minefewcoins]

The PSU was sent at shop to check it out and it was loaded at almost max load and working 12+h without restarts. I can't test it because I don't have second PSU powerful enough.

it's a right decision, on a GPUs Rig troubleshooting, the physical check is important, check devices position, temperature, etc.

My rigs have been running at 88 degrees for 2 years now, yet you guys are thinking that 80+ degrees is a lot.

Around 85 degrees is just fine for GPU's in my opinion.

That's nothing with mine. I run two days my Gigabyte Windforce 290X with the temperature between 91-92, I stop it cause my PSU was burning. Now, that card run normally, under 85 Lol. Yeah, it's normal if I compare it with the temperature before.

Wait did you say that bad psu caused GPUs to run 6-7C hotter?

[Geraldo]

Wait did you say that bad psu caused GPUs to run 6-7C hotter?

Yes, for example, your PSUs didn't run on correct voltage, your cards would start drawing more power to attempt to bring the voltage back up. It will make your PSUs overloaded and also make your card working harder that would produce more heat and can make a random restart on your systems.

[minefewcoins]

Wait did you say that bad psu caused GPUs to run 6-7C hotter?

Yes, for example, your PSUs didn't run on correct voltage, your cards would start drawing more power to attempt to bring the voltage back up. It will make your PSUs overloaded and also make your card working harder that would produce more heat and can make a random restart on your systems.

Looks like a legit cause. Thank you for sharing it with me

[Geraldo]

Wait did you say that bad psu caused GPUs to run 6-7C hotter?

Yes, for example, your PSUs didn't run on correct voltage, your cards would start drawing more power to attempt to bring the voltage back up. It will make your PSUs overloaded and also make your card working harder that would produce more heat and can make a random restart on your systems.

Looks like a legit cause. Thank you for sharing it with me

Yeah, of course, we are happy to help other members here. Report here what you've done after. It would help other members with the same problems.

[kissmarx]

Hi op. Are you still mining? Was the problem on this rig been already solved? How is it doing now? I'm interested to know because I have similar problem now. And I'm trying to make my rig dual psu to ease the load of the first psu and to see if it will be solved.

[Kalpesh2872]

Can you tell someone if can i invest in mining? If is it real?

[Coinluminary]

Check cards one by one there would be something wrong with single cards and that's why it is affecting the whole rig.

[kissmarx]

Sometimes it is one of the cards causing the problem. That's why I agree that you test your cards one by one.

Sometime what's causing the problem are the risers. Sometimes when the PSU is overloaded than what is capable of, can also cause the problem.

It is important to exhaust all the possible hardware problems. Then there is also the software problem to exhaust.