Unable to determine the device handle for GPU0000:61:00.0: Unknown Error

Gaming and Visualization Technologies General Topics and Other SDKs
1

My rig keeps having issues with a particular video card not being recognized when running calculations, and I’ve read a lot of posts in the forums that seem to be temperature or power supply related.

image
image1920×1177 350 KB

Here’s the driver log I printed, the error serial number is 79, it seems to still be temperature or power supply

2

image
image1924×1316 334 KB

But before this device was lost, its problems did not reach the threshold for forced shutdown.
Current system is: ubuntu 22.04
nvidia version: 550.90
cuda : 12.40
Graphics card model: 4090 * 4

The test has persistent mode enabled, but still the same error.