PCs throttle performance when they get too hot. Games stutter. Apps crash. Fans whine constantly. Checking your CPU and GPU temperatures takes 2 minutes with the right tools and tells you exactly what's going on.
Here are the free tools and what numbers actually matter.
Check CPU temperature
Download Core Temp (free, no install needed). Run it. The main window shows each CPU core's current temperature in real time.
Look at the Max column – that's the hottest your CPU has reached since you opened the app. The Tj. Max value is your CPU's thermal limit. Anything within 10°C of that is concerning.
Check GPU temperature
For graphics card temps, use GPU-Z or HWMonitor. Both free. GPU-Z is simpler, HWMonitor shows more details.
Open the app and run a game or stress test in the background. Watch the GPU temperature. The reading updates in real time and shows you what happens when the card is actually working.
What temperatures are safe
| Component | Idle | Under load | Danger zone |
|---|---|---|---|
| CPU (most modern) | 30-50°C | 60-80°C | Above 90°C |
| GPU (NVIDIA) | 30-50°C | 65-85°C | Above 95°C |
| GPU (AMD) | 30-50°C | 70-90°C | Above 100°C |
| NVMe SSD | 40-60°C | 65-75°C | Above 85°C |
| Motherboard | 30-40°C | 40-50°C | Above 70°C |
If your temps land in the "Danger zone" column, your PC will throttle performance to protect itself. You'll feel it as stutters or slowdowns. Time to address cooling.
Stress test to see worst-case temps
If you want to know how bad it gets under load, run a stress test:
- Cinebench R23 – quick 10-minute CPU stress test
- Prime95 – long-form CPU stress, finds the absolute hottest temps
- FurMark – GPU stress test, runs the card at 100%
- 3DMark – more realistic gaming load
Run one for 10-15 minutes while watching temperatures. The peak you see is essentially the worst case you'll hit during real use. If peak is safe, you're fine.
Use HWMonitor for everything at once
HWMonitor (or its prettier sibling HWInfo64) shows every sensor in your PC in one window. CPU, GPU, motherboard, SSD, fan speeds, voltages. Pretty much everything.
Download free from cpuid.com. Open it. Browse the tree to find the readings you want. The Min, Value, Max columns show recent history.
If your temps are too high – cooling fixes
Hot PCs usually have one of these problems:
- Dust buildup – clean fans and heatsinks with compressed air
- Old thermal paste – replace after 3-5 years on CPU
- Poor case airflow – add more intake or exhaust fans
- Bad fan curves – configure fan ramp in BIOS or with software
- Failing fan – listen for grinding, replace if needed
Dust is the most common one. A heavily dusty PC can run 15-20°C hotter than a clean one. Quick compressed air cleaning fixes most of it.
Build a temperature log
If temps fluctuate or you're trying to track down a heat issue, HWInfo64 can log to a CSV file. Run it during a session, save the log, open in Excel to graph the temps over time.
You'll see spike patterns that point to which apps or activities cause heat problems. Sometimes a misbehaving background process is the cause, not the workload you expected.
Laptop-specific tips
Laptops run hotter than desktops because of small cases. Strategies that help:
- Use on a hard surface, not a bed or lap (fabric blocks intake vents)
- Get a laptop cooling pad with fans, $20-40
- Repaste the CPU/GPU thermal compound (worth doing every 2 years)
- Undervolt the CPU using Intel XTU or Ryzen Master
Cooling pads alone often drop temps 5-10°C. Cheap fix that buys a lot of headroom.
What temps are you seeing? Drop your CPU and GPU model plus their idle and load temps. I'll tell you if it's normal for that combo.