10x Vega 56 BSOD

[wharmus]

Hello guys , today I encountered a weird problem.
Power went down to the site where I had like 20 rigs , all of them came back when power came back on and all of them started mining again without any problems.
All of them except my 10 x vega 56 rig , that has been mining for 2 weeks.
Power went down adn when it came back up , I started the miner and it would BSOD after some time.
I m using adrenalin 18.3.4 . I removed the drivers , I put the latest , I started mining , the same problem.
I put back 18.3.4 , the same problem. I dont knwo what to do , after some time , BSOD.
I m mining CryptoNight with cast xmr , I m overclocking with OverNDrive Tool, by applying the settings and then diabling/enabling the gpus.
My settings are :


[Profile_2]
Name=Vega56
GPU_P0=852;900;0
GPU_P1=991;900;0
GPU_P2=1084;900;0
GPU_P3=1138;900;0
GPU_P4=1150;900;0
GPU_P5=1202;900;0
GPU_P6=1212;900;0
GPU_P7=1375;900
Mem_P0=167;900;0
Mem_P1=500;900;0
Mem_P2=700;900;0
Mem_P3=920;900
Fan_Min=2600
Fan_Max=4900
Fan_Target=60
Fan_Acoustic=2300
Power_Temp=75
Power_Target=-15

If anyone has any idea what to do I'm all ears.
thanks

[1714052658]

1714052658

[Metroid]

overheating somewhere perhaps, check everything concerning overheating, track down issues can be very time consuming and mind blogging. it can be anything really.

[yrk1957]

BSOD code may help.

[wharmus]

its not overheating. and it was working perfectly until the power down toady.
even now im reinstalling the drivers again , it gave bsod after 1 hour.


The computer has rebooted from a bugcheck.  The bugcheck was: 0x00000116 (0xffffd48a4ea880f0, 0xfffff80af3d0fc7c, 0xffffffffc0000001, 0x0000000000000003). A dump was saved in: C:\Windows\MEMORY.DMP. Report Id: 0f4ed64f-a785-4cac-bd28-524cc1a6dd44.

[yrk1957]

The computer has rebooted from a bugcheck.  The bugcheck was: 0x00000116 (0xffffd48a4ea880f0, 0xfffff80af3d0fc7c, 0xffffffffc0000001, 0x0000000000000003). A dump was saved in: C:\Windows\MEMORY.DMP. Report Id: 0f4ed64f-a785-4cac-bd28-524cc1a6dd44.

Maybe power related. Start by reducing load on each PSU by removing one card each off each PSU.

Edit: Did you have soft power play tables applied? Maybe you need to re-apply?

[wharmus]

I have the bad habbit of using strong PSU's for my rigs.
For example this one has 6 x Vega56 on Ax1500i and 4 x Vega 56 on Hx1200i , Corsair.
I m measuring power from the wall , its always been 900W plus minus 20 watts.
It's not power 100%. Im thinking maybe some riser got screwed during the power down.

[yrk1957]

I have the bad habbit of using strong PSU's for my rigs.
For example this one has 6 x Vega56 on Ax1500i and 4 x Vega 56 on Hx1200i , Corsair.
I m measuring power from the wall , its always been 900W plus minus 20 watts.
It's not power 100%. Im thinking maybe some riser got screwed during the power down.

Well, then elimination is the only option.

Edit: One more thing about Vega56 and power. I had 6x 56s drawing 7a/240v, that's 280w per card. So that Hx1200i maybe running pretty close with 4x 56s.

[wharmus]

Yeah thanks for the help.

[wharmus]

UPDATE : I tried running Claymore Cryptonote Miner , it doesnt make the same hashes as cast , but I got an error on GPU 9 .
I checked the settings everything is correct , so I started cast-xmr without GPU9.
Now to see if it works and doesnt bluescreen anymore. Is there a chance that the Gpu got affected by the pwoerdown somehow?
Thanks

[Vann]

Yes. How many PSU's are powering the VGA power connectors on that card?

[wharmus]

Update , even with GPU9 disabled , it still went down , BSOD.
What do you mean VGA connectors? U mean the PCI-e Cables?
there's 2 PCI-e connectors on each card.

I dont power 1 card from 2 PSU's.
I have 6 cards on one PSU , 4 cards on the other.

[Vann]

Yes. Only one PSU should power both VGA power connectors. If you use two different PSU's, they could go out of phase and damage the GPU. Disconnect that card from the PCI-E slot on the motherboard and see if you can boot without out it.

[wharmus]

I can boot. The BSOD happens when I start mining.
After some time, can be 1 hour , can be 5 min.

[Vann]

A bad GPU can make the whole system unstable, which is why you want to disconnect it from the motherboard so you can isolate the problem. You could also try that GPU in another system by itself.

[Specialist1]

For the original poster... This is exactly what I recommend you do also... pull off the riser and unplug the VGA cables to the card in question

A bad GPU can make the whole system unstable, which is why you want to disconnect it from the motherboard so you can isolate the problem. You could also try that GPU in another system by itself.

[szafa]

My bsod had happend when psu dying and finally dead,new psu and bsod resolved.

[Specialist1]

Interesting... same size PSU?

My bsod had happend when psu dying and finally dead,new psu and bsod resolved.

[wharmus]

no, 1 x ax1500i , 1 x hx1200i .
i think i have one faulty gpu , somehow i removed it and it seems that the BSOD stopped.
I ve been mining for 15h now , no BSOD yet.

[huntingthesnark]

As somebody else said, try swapping the riser on that 'faulty' gpu - I had a power outage that killed a 5vega rig, took bloody ages to troubleshoot.

New riser got it back up eventually - had been running fine for months previous, clearly the riser took a hit somehow when the power when out/back on again.

[Bare]

no, 1 x ax1500i , 1 x hx1200i .
i think i have one faulty gpu , somehow i removed it and it seems that the BSOD stopped.
I ve been mining for 15h now , no BSOD yet.

As somebody else said, try swapping the riser on that 'faulty' gpu - I had a power outage that killed a 5vega rig, took bloody ages to troubleshoot.

New riser got it back up eventually - had been running fine for months previous, clearly the riser took a hit somehow when the power when out/back on again.

Yeah, let's hope your riser is faulty and not your gpu, as people have already posted, you'll want to try the gpu in question with another riser,
if the same problem persists, test your gpu in a different rig and if it doesn't work there too, well, your gpu is probably busted, but if it works then I might be suspecting the pcie slot on your rig's motherboard might have a problem...

in any way, good luck!