[Updated] Watchdog DPC Violation Error with Multiple NVIDIA GPUs

Introduction

We have seen an issue in our Support department with a specific hardware combination: the Asus X99E-10G WS motherboard and triple or quad NVIDIA GPU configurations. When running a workload that is heavy on both the CPU and GPUs, systems with this hardware can crash and report DPC_WATCHDOG_VIOLATION “blue screen of death” (BSOD).

Since the initial discovery of the issue, we have determined that the issue affects any Nvidia multi-GPU systems with motherboards featuring PLX chips (such as the Asus X99E-10G WS)  and Nvidia driver versions higher than 382.53 (released in 2017).

Solution

As it isn’t realistic to continue running a graphics driver from 2017, we now recommend a solution that involves instructing the graphics driver to use “Message Signaled-based Interrupts-mode” or “MSI-mode” for short. By default, the drivers use the older “Line-based” mode, which leads to these BSODs when paired with multiple Nvidia GPUs and motherboards using PLX chips.

Please navigate HERE and download the MSI utility onto the affected system. After downloading and extracting the files, run the executable as an administrator, and then perform the adjustment per the included “README” file.

For more information about Line-based vs. MSI interrupts, we recommend checking out this discussion.

Conclusion

If you are a Puget Systems owner and have any questions or prefer our Puget Systems Support staff perform the adjustment, please don’t hesitate to reach out; we are here to help!