GPUAlert Enterprise is a lightweight standalone Linux monitoring tool for NVIDIA and AMD GPUs. It automatically detects the GPU vendor and monitors GPU utilization, VRAM usage, temperature, power, hardware health, system memory, and active GPU processes.
Alerts are sent after a configurable number of consecutive threshold violations. Matching recovery notifications are sent when conditions return to normal, reducing alert noise.
Notification channels include Email, Telegram, Microsoft Teams, Slack, Discord, LINE, PagerDuty, SMS, and generic webhooks. GPUAlert operates fully offline and does not require a monitoring server.
This download runs in Trial Mode for evaluation. Paid offline node licenses and annual site licenses are available.
Documentation: https://www.smartbooth.com/gpualert.htm
Licenses: https://www.smartbooth.com/gpualert_lic_purchase.htm
Features
- gpu-monitoring
- nvidia-gpu
- amd-gpu
- vram-monitoring
- temperature-monitoring
- email-alerts
- telegram-alerts
- offline-monitoring
