How to Set Up HWiNFO Sensor Alerts
HWiNFO sensor alerts are user-defined alarms that watch a chosen sensor (such as CPU temperature) and fire an action — a sound, popup, log entry, or launched program — the moment a reading crosses your threshold. For gamers, they turn passive monitoring into an active early-warning system for overheating and throttling that Windows never warns you about.
HWiNFO ships with no pre-defined alerts — every alert is configured manually, per sensor. This guide walks you through enabling alerting, picking safe thresholds, and choosing actions that actually reach you mid-game.
Already have a log? Skip the manual setup — upload your HWiNFO CSV to our free analyzer and it flags throttling, power limits and bottlenecks for you in seconds.
Analyze my log freeWhat HWiNFO Alerting Does
Alerting is a per-sensor feature: you pick one value, enable alerting on it, set an above/below threshold, and attach an action. The best use case for gamers is catching silent thermal throttling and power/clock drops — these happen at the hardware level and produce no OS notification, so your FPS quietly tanks with no warning.
Hard requirement: HWiNFO and its Sensors window must be running for alerts to fire. If HWiNFO is closed, alerts do nothing. Launch it in Sensors-only mode at Windows startup so your alerts are always on. New to the sensors? Start with how to read HWiNFO indicators for gaming.
Step-by-Step: Create a Sensor Alert
- Open the Sensors window. Launch HWiNFO and click Sensors (the gauge icon). Check Sensors-only on the startup screen for low overhead while gaming.
- Find the sensor you want to watch — e.g. CPU Package, GPU Temperature, GPU Hot Spot, Thermal Throttling (YES/NO), or fan RPM.
- Open that sensor's settings. Double-click the sensor row (or right-click → settings), then go to the Alerts tab.
- Enable alerting. Tick Enable Alerting for the value.
- Set the threshold. Define the value above which (for temps/power) or below which (for clocks/fan RPM) the alarm triggers. Note: "above" triggers are more reliable than "below" triggers in current builds.
- Set "Samples to trigger alarm." How many consecutive over-threshold readings are required before firing. At 1-second polling,
5means the condition must hold for 5 seconds — this filters out harmless momentary spikes. - Set "Notification distance." The minimum time between repeat alarms while the condition persists (e.g.
30= at most one alert every 30 seconds). Optionally enable trigger only once. - Choose the action(s). Combine any of: play a sound, show a warning window, turn the value red, log to a text file, or launch an application.
- Repeat for each critical sensor, then leave HWiNFO/Sensors running in the background during your session.
Recommended Alert Thresholds
Use these as a starting point and tune to your specific CPU/GPU — limits vary by silicon (Ryzen targets ~95°C; some Intel parts 100°C). For longevity, aim to stay under 80°C sustained.
| Sensor (HWiNFO label) | Direction | Threshold | Samples | Why it matters |
|---|---|---|---|---|
| CPU Package / Tctl-Tdie | Above | 90°C | 5 | Approaching throttle; ~95–100°C is the hardware limit on most CPUs |
| GPU Temperature | Above | 83°C | 5 | Average die temp; 85°C+ risks throttling on many GPUs |
| GPU Hot Spot | Above | 100°C | 3 | Hottest point on the die, ~10–20°C above GPU Temp — the real danger number |
| CPU Thermal Throttling | = YES | 1 / YES | 1 | Direct confirmation the CPU is dropping clocks from heat |
| CPU Power Limit Throttling | = YES | 1 / YES | 1 | Clocks capped by power limit (PL1/PL2 or PPT) |
| GPU Performance Cap – Thermal | = YES | 1 / YES | 1 | Explains sudden GPU usage/clock drops |
| GPU Effective Clock | Below | ~30% under normal | 5 | Catches clock-drop/throttling not shown by temperature alone |
| VRM / VR MOS (Motherboard) | Above | 100°C | 5 | Hot VRMs throttle the CPU; flags poor case airflow |
| Fan RPM (CPU/GPU/case) | Below | ~300 RPM | 3 | Detects a stalled or failed fan before temps spike |
For the boolean flags (Throttling / Performance Cap), HWiNFO reports YES/NO — use an "above 0" trigger with 1 sample for an instant alert. "Effective Clock" thresholds must be personalized: log a baseline gaming session first to learn your normal clocks.
Which Alert Action Should You Use?
Sound
Audible cue while you're focused on the game — the most practical choice for fullscreen play.
Warning window
Clearest, but pauses sensor updates until dismissed — an ignored popup can block later alarms.
Mark value red
A passive visual flag in the Sensors window — fine as a secondary indicator.
Log to text file
Best for after-the-fact diagnosis: records which alarm fired and when. Then run deeper HWiNFO log analysis.
Launch application
Most flexible/reliable — call a small notifier or script. Recommended for fullscreen gaming where popups are hidden.
Polling matters: the default global polling period is 2000ms, and alerts can only react as fast as sensors are polled. For gaming diagnostics, lower it to 500–1000ms so your "samples to trigger" math gives a meaningful sub-5-second response.
Common Pitfalls
- HWiNFO/Sensors closed = no alerts. The single most common failure. Keep it running (Sensors-only at startup).
- Popup window freezes monitoring. A displayed warning window halts all sensor updates until closed, which can suppress the next alert. Prefer sound or a launched app for fullscreen use.
- "Below" triggers are flaky. Above-value triggers are reliable; below-value triggers (clock-drop, fan-RPM) have known quirks — test them before trusting them.
- No samples buffer = false alarms. Setting samples-to-trigger to 1 on a noisy temperature sensor fires on harmless spikes. Use a few seconds' worth of samples for temps.
- Wrong threshold per silicon. A 90°C alert on a chip that legitimately runs to 95°C creates alert fatigue. Tune to your hardware.
- Minimized-to-tray quirk. Some power/fan sensors save alert state while minimized and fire all at once when reopened — verify your critical alerts trigger live.
Frequently Asked Questions
Does HWiNFO have to stay open for sensor alerts to work?
Yes. HWiNFO and its Sensors window must be running for any alert to fire. If the program is closed, alerts are inactive. The lightweight Sensors-only mode, set to launch at Windows startup, is the recommended way to keep alerts always-on.
What CPU and GPU temperatures should I set alerts at for gaming?
A practical starting point is 90°C for CPU Package and 83°C for GPU Temperature, plus a 100°C alert on GPU Hot Spot. Adjust to your hardware — many modern CPUs only throttle near 95–100°C — and aim to stay under 80°C sustained for longevity.
Why didn't my HWiNFO alert trigger even though the temperature spiked?
Common causes: HWiNFO or the Sensors window wasn't running, 'Samples to trigger alarm' was set too high so a brief spike didn't last long enough, the polling period was too slow (default 2000ms), or you used a 'below' trigger, which is less reliable than 'above'.
How do I get an audible alarm instead of just a red value?
In the sensor's Alerts settings, enable the 'Play sound' action (and optionally 'Display warning window'). For fullscreen games where popups are hidden, attach a 'launched application' or sound action rather than relying on the visual red-value indicator.
How can I alert on thermal throttling specifically?
Set alerts on the boolean flags directly: Thermal Throttling and Power Limit Throttling for the CPU, and Performance Cap – Thermal for the GPU. These report YES/NO, so use an 'above 0 / YES' trigger with 1 sample for an instant warning the moment clocks are being cut.
Got an alert and a log? Upload your HWiNFO CSV and let the analyzer pinpoint the exact cause of your throttling or FPS drops.
Upload for Analysis