Make sure NVIDIA Fabric Manager is running and has the same version number as the GPU drivers.
Why does my Droplet have high CPU or RAM usage?
Validated on 26 Feb 2021 • Last edited on 18 Mar 2026
High RAM or CPU usage is normally the result of applications or kernel processes on the Droplet itself. While we do monitor CPU and RAM usage on the physical host the Droplet resides on (the hypervisor), the usage on the physical host itself doesn’t translate to the CPU usage within the Droplet itself.
If high RAM or CPU usage is causing performance issues on your Droplet, you can log in to the Droplet and monitor which processes are potentially abusing the Droplet’s resources and stop them if necessary. If your Droplet chronically experiences high RAM or CPU usage under normal workloads, you can resize the Droplet to a larger plan at any time.
Related Topics
There may be an issue with the autoscale pool or Droplet configuration, the VPC network’s size, or resource limits on the team or datacenter.
Health checks often fail due to firewalls or misconfigured backend server software.