Monitoring
Introduction
Section titled “Introduction”This chapter defines what to monitor, what abnormal readings may indicate, and how to investigate without inventing exact limits.
Purpose
Section titled “Purpose”Use monitoring to identify cooling, power, driver, memory, storage, and airflow problems early.
Estimated Time
Section titled “Estimated Time”20-40 minutes for setup; 5-10 minutes for routine checks.
Difficulty
Section titled “Difficulty”Moderate.
Required Tools
Section titled “Required Tools”- HWiNFO64 or equivalent monitoring software.
- AMD Adrenalin for GPU telemetry.
- Samsung Magician for SSD health.
- Baseline log from Benchmark Baseline.
Warnings
Section titled “Warnings”- Do not treat one spike as a failure without workload context.
- Do not compare readings from different tools unless sensor names match.
- Do not run stress tests unattended.
- Do not use unverified internet temperature targets as hard limits.
Monitored Values
Section titled “Monitored Values”| Value | Normal use expectation | Warning threshold | What abnormal values may indicate | Investigation sequence |
|---|---|---|---|---|
| CPU temperature | Idle and gaming readings should be consistent with the original baseline and temperature reference. | Investigate sustained readings materially above baseline in similar room conditions, rapid temperature rise, or thermal throttling. | Cooler mount, pump/fan reading, radiator airflow, dust, background load, BIOS fan curve. | Check workload, pump/fan readings, radiator fans, dust, then cooler mount. |
| GPU temperature | Gaming readings should be stable for the same game/settings. | Investigate sustained rise versus baseline, driver timeouts, artifacts, or fan behaviour changes. | Blocked bottom/side intake, GPU fan curve, driver issue, dust, cable obstruction. | Check fans, airflow, driver version, game settings, and case filters. |
| SSD temperature | Idle/light use should be stable; sustained transfers run warmer. | Investigate high idle temperature, thermal throttling, or large temperature rise during normal use. | Heatsink contact, poor airflow, heavy background writes, low free space. | Check Samsung Magician, drive free space, background activity, heatsink seating. |
| Motherboard temperature | Should remain stable relative to baseline. | Investigate if it rises with no workload or after airflow changes. | Case airflow, cable obstruction, fan failure, sensor mismatch. | Check case fan operation and recent fan curve changes. |
| CPU package power | Should scale with workload. | Investigate high power at idle or unexpectedly low power under load. | Background process, power plan, BIOS setting, thermal limit. | Check Task Manager, power plan, CPU clocks, temperatures. |
| GPU power | Should scale with game or benchmark load. | Investigate if high at idle or unstable under load. | Driver setting, high-refresh desktop behaviour, overlay, power cable issue. | Check refresh rate, Radeon settings, background GPU use, PCIe power seating. |
| Clock speeds | Should boost and idle according to load. | Investigate clocks stuck low under load or stuck high at idle with high power. | Thermal throttling, power limits, background load, driver setting. | Check temperatures, utilisation, power, Windows power plan. |
| Memory speed | Should match default or EXPO state recorded in BIOS/Windows. | Investigate if speed changes unexpectedly or instability appears after EXPO. | BIOS reset, failed training, unstable EXPO, wrong slots. | Check BIOS EXPO state, Task Manager speed, MemTest86 if unstable. |
| Fan speeds | Should follow BIOS or FanControl curves. | Investigate missing fan reading, unexpected 0 RPM under load, or new noise. | Header issue, fan stop mode, cable obstruction, fan curve conflict. | Check BIOS/FanControl, physical fan movement, cables. |
| Voltages | Should be reviewed mainly for changes from baseline or obvious sensor faults. | Investigate sudden abnormal readings, instability, or values flagged by monitoring software. | Sensor misread, PSU issue, motherboard reading difference, overclock setting. | Cross-check tool readings, restore defaults, inspect power cabling. |
Step-by-Step Instructions
Section titled “Step-by-Step Instructions”- Open HWiNFO64 sensors-only mode.
- Confirm CPU, GPU, SSD, motherboard, fan, and memory sensors are visible.
- Wait 10 minutes at idle after startup.
- Record idle readings.
- Run a short controlled load from Benchmark Baseline.
- Record load readings and fan behaviour.
- Save the baseline privately.
- Compare future readings against the same workload, room context, driver version, and fan profile.
Verification Checklist
Section titled “Verification Checklist”- CPU temperature is visible.
- GPU temperature and utilisation are visible.
- SSD temperature and health are visible.
- Fan speeds are visible where headers expose them.
- Memory speed is recorded.
- Baseline includes idle and controlled load readings.
- Monitoring tool version is recorded.
Common Mistakes
Section titled “Common Mistakes”- Comparing idle after boot with idle after a game session.
- Using different fan profiles for baseline and retest.
- Ignoring room temperature.
- Treating every voltage sensor label as equally reliable.
- Running multiple sensor tools and confusing duplicate readings.
Expected Result
Section titled “Expected Result”Monitoring provides a repeatable view of thermal, power, clock, fan, storage, and memory behaviour.
Next Chapter
Section titled “Next Chapter”Continue to FanControl Configuration.