Add support for GPU metrics gathering/monitoring to Azure Portal Monitor tool
yes, it is important with gpu computing.
we use 4 V100, and recently I understood that we load them under 50%. Would be better to watsh it in monitoring
Ben Huntley commented
This would be very valuable to our help our team maximize our Azure utilization.
Oleg Losinets commented
Please help us to do our work more efficiently. We need GPU monitoring asap.
Kris Zentner commented
Would be wonderfully useful to get an idea of utilization on NV systems for our HPC clusters and Data Science VMs. Getting utilization can help us right size our Azure spend!
Chris Erdman commented
Monitoring of GPU utilization and FPGA utilization is important to understanding how effectively we are using our Azure resources.
Robert Eberl commented
Having this monitor can help our customers "right-size" the costs they incur for the work they need to accomplish with GPUs in Azure, I expect that can be very valuable to our sales teams!
Stephen Dahl commented
Can you guys help with using DCGMI or another tools to push to metrics to help us see what usage of GPU's in our cluster?
GPU utilization is very important for our team as we need to make sure people are using GPUs fairly. Please enable this feature as soon as possible
Jim Jernigan commented
GPU Utilization (% time, RAM) is a critical tool for us to determine if resources are being used effectively, both at the job level and the resource allocation level. Please enable it as soon as possible.
This is important to monitor if users are actually using the GPUs or to recommend using CPU only machines. This can represent cost svgs