RE-Implemented NVIDIA Energy capture via C #1167
Conversation
![greptile-apps[bot]](https://avatars.githubusercontent.com/in/867647?s=60&v=4){kind=small}
There was a problem hiding this comment.
4 file(s) reviewed, 3 comment(s)
Edit PR Review Bot Settings | Greptile
| {"help", no_argument, NULL, 'h'}, | ||
| {"interval", no_argument, NULL, 'i'}, |
There was a problem hiding this comment.
logic: interval option is marked as no_argument but requires an argument in getopt_long
| {"help", no_argument, NULL, 'h'}, | |
| {"interval", no_argument, NULL, 'i'}, | |
| {"help", no_argument, NULL, 'h'}, | |
| {"interval", required_argument, NULL, 'i'}, |
|
@ribalba Can you please check what the correct install command is for the libraries under Fedora. ChatGPT suggested: sudo dnf install cuda-nvml-dev It also mentioned it is not in the distributions repos and you need to add: sudo dnf config-manager --add-repo=https://developer.download.nvidia.com/compute/cuda/repos/fedora$(rpm -E %fedora)/x86_64/cuda-fedora.repo |
|
Eco CI Output:
🌳 CO2 Data: Total cost of whole PR so far: |
We were experiencing some sampling-rate issues with the
nvidia-smiimplementation where the sampling jitter was too high.This is a re-implementation in C which is still minimally slower than our other providers, but quite performant so we can achieve sampling < 100 ms
Greptile Summary
Re-implemented NVIDIA GPU power monitoring from Bash to C using NVML library, significantly improving sampling rate consistency and enabling sub-100ms measurement intervals.
source.cwith direct NVML library integration for precise GPU power measurementsmetric-provider-nvidia-smi-wrapper.shshell script to eliminate sampling jitter issuesMakefilewith -O3 optimization and NVML library linkage for performanceprovider.pyto support card model identification and handle the new C-based metrics formatsource.cto prevent thread state persistence