bstollenvidia

Results 6 comments of bstollenvidia

Errors in DcgmReaderExample.py

I confirmed this is still present in 3.1.6 https://github.com/NVIDIA/DCGM/blob/4aedfaae1f7c8480e46b8c835ddd5afbd00d57be/testing/python3/DcgmReader.py#L400

python binding process smutilization always return 2147483632

A couple debugging steps: Can you share the output from the following commands: 1. dcgmi --version 2. dcgmi discovery -l

NVVS (part of NVIDIA DCGM) has replaced nv-healthmon. NHC will fail on new GPUs w/o code mods

All DCGM packages and docs can be obtained here: https://developer.nvidia.com/data-center-gpu-manager-dcgm

Replacement for NVIDIA_HEALTHMON check?

NVVS (soon to be called DCGM GPU Diagnostic) is part of the DCGM package and can be obtained here: https://developer.nvidia.com/data-center-gpu-manager-dcgm Note that the link to the NVVS user guide is...

Support for A30/40 and L30 GPUs?

DCGM supports profiling metrics on all Volta and newer compute GPUs, including Ampere and Ada ones.

v4.1.1 failed to load profiling module

Can you please retry with DCGM 4.2.3? There were several fixes to how we load the profiling module.