strix-halo-monitoring

Monitor CPU/GPU temps, power draw, and clock speeds on Strix Halo during inference workloads.

Updated Apr 19, 2026
One-click install
npx skills add https://github.com/crycriM/hermes-skills --skill strix-halo-monitoring
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: strix-halo-monitoring
Source: https://github.com/crycriM/hermes-skills/tree/main/mlops/strix-halo-monitoring
Command: npx skills add https://github.com/crycriM/hermes-skills --skill strix-halo-monitoring

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires bc, and includes scripts (resource) components.

What problem does it solve?

Real-time monitoring of CPU/GPU temps, power draw, and clock speeds on the Strix Halo APU during inference workloads to help prevent overcurrent events and thermal throttling.

Core Features & Use Cases

  • Hardware sensor map (hwmon) references sensors for temperature, power, and clock readings to provide a comprehensive view during inference.
  • Monitoring script (monitor.sh) logs CSV data and prints live readings, enabling quick diagnostics and long-run analysis.
  • Power safety guidelines describe BIOS mode considerations and circuit safety to avoid breaker trips on shared circuits during sustained workloads.

Quick Start

Launch the Strix Halo monitor script to start real-time thermal and power logging during an inference workload.

Frequently Asked Questions about strix-halo-monitoring

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I monitor CPU and GPU temperatures on Strix Halo during inference?

You can monitor Strix Halo temperatures during inference by running a dedicated hardware sensor map script that reads hwmon data. It provides live thermal readings to help prevent thermal throttling in real-time.

What is the best way to track power draw on Strix Halo to avoid tripping a breaker?

Tracking power draw on Strix Halo involves running a monitoring script that logs real-time power consumption data. It includes power safety guidelines and BIOS mode considerations to prevent overcurrent events on shared circuits.

How does the Strix Halo monitoring script log sensor data for long-running workloads?

The Strix Halo monitoring script logs sensor data by writing hardware readings into a CSV file. This enables long-run analysis of clock speeds, temperatures, and power draw during sustained inference workloads.

Does the Strix Halo monitor require any specific dependencies to read hardware sensors?

Yes, the Strix Halo hardware monitor requires the bc command-line calculator dependency to process sensor readings. The monitoring script uses bc to parse and calculate live thermal and power data from hwmon.

Can I use this monitoring script for long-running inference on a Bosgame M5?

Yes, the Strix Halo monitoring script is specifically designed for long-running inference scenarios on Bosgame M5 hardware. It tracks real-time thermal and power metrics to ensure safe operation in production-like settings.

Why should I monitor clock speeds and power during Strix Halo inference workloads?

Monitoring clock speeds and power during Strix Halo inference workloads is crucial to prevent thermal throttling and overcurrent events. Real-time hardware sensor tracking helps maintain safe operating temperatures and avoid breaker trips.