DPC WATCHDOG VIOLATION on Supermicro 4028GR-TRT2 Tweet

[coilbook]

Hi,
We bought SYS-4028GR-TRT2 for octane GPU rendering as a rendering slave. We have 10 GPUs 8 titan x pascal and 2 2080ti hybrids. We keep getting BSOD with dpc watchdog violation while it is rending. I attached the link to wetransfer where you can download dump file. We are not sure why. We have the latest nvidia drivers. Maybe it is related to nvidia drivers.

Can anyone read through DMP file or if anyone has supermicro knows what it can be?

We have another regular PC and it is not getting this error. Thank you,
Dump file link: https://wetransfer.com/downloads/bd6b98 ... 616/870ea4

Thank you

[coilbook]

has anyone got a bad RTX 2080ti? Can this be a reason

[xanderjm]

1. Download the MSI Utility v2.0 zip file here: https://github.com/CHEF-KOCH/MSI-utility/releases
2. Open the zip file and then extract "MSI_util_v2.exe" to your local system.
3. Right-click the "MSI_util_v2.exe" and Run as administrator.
4. Enable MSI for each GPU listed.
5. Restart the system.

This is the advice i had from nvidia as a workaround untill they had a fix, they said they are aware of the problem and i beleive its very specific to some motherboards.

[coilbook]

1. Download the MSI Utility v2.0 zip file here: https://github.com/CHEF-KOCH/MSI-utility/releases
2. Open the zip file and then extract "MSI_util_v2.exe" to your local system.
3. Right-click the "MSI_util_v2.exe" and Run as administrator.
4. Enable MSI for each GPU listed.
5. Restart the system.

This is the advice i had from nvidia as a workaround untill they had a fix, they said they are aware of the problem and i beleive its very specific to some motherboards.

Thank you!

[coilbook]

1. Download the MSI Utility v2.0 zip file here: https://github.com/CHEF-KOCH/MSI-utility/releases
2. Open the zip file and then extract "MSI_util_v2.exe" to your local system.
3. Right-click the "MSI_util_v2.exe" and Run as administrator.
4. Enable MSI for each GPU listed.
5. Restart the system.

This is the advice i had from nvidia as a workaround untill they had a fix, they said they are aware of the problem and i beleive its very specific to some motherboards.

By the way could you tell me what the tool does? Does it run on background along with drivers and corrects errors?
Thank you

[coilbook]

Ok. It looks like msi utilities does help. But now I get this strange crash where afterburner gpu graph would just stop moving and pc gets very slow. 15-20 seconds between each button pressed and it must be rebooted. I’ve never had this problem before. This server has its own vga output driver and it holds 10 gpus for calculations only. I wonder if MSI Utilities software prevents pc from bsod but Nvidia drivers still crash and pc become very slow. Has anyone had this problem before?

[paride4331]

Hi coilbook,
that tool fixs PLX issue, I think your motherboard has not PLX, so it does not solve your issue.
My advice is to ask supermicro service assistance.
Regards
Paride

[coilbook]

Hi coilbook,
that tool fixs PLX issue, I think your motherboard has not PLX, so it does not solve your issue.
My advice is to ask supermicro service assistance.
Regards
Paride

Thanks Paride,
they told me to try windows server 2016. They don't know and I never had this problem before. It would render for 2 hours or so and then stop and get very slow even though no errors about crashes or nvidia drivers failing.
Even master will see it as slave is connected but when it's time to render it won't render until it is rebooted. It has two xeon Intel Xeon E5-2620 V4 Broadwell-EP 2.1 GHz.
When using out of core it gets even worse I wonder if CPUs cannot keep up sending data to vram.

Thanks

[paride4331]

Hi coilbook,
try asking to smicha, I think it know quite well those motherboards
Regards
Paride

[coilbook]

just in case if anyone has same problems a lot of BSOD was related to msi afterburner and custom fan profile. Not sure why.