Short answer: how hot is too hot for a computer?
It depends on the part and the load, but as a rough guide:
- CPU under gaming load: about 60–85°C is typical and comfortable.
- CPU at 90°C and above: worth a closer look. On many Ryzen chips, 90–95°C under heavy load is by design, so the number alone does not prove a fault.
- Sustained readings within about 5–10°C of the chip’s TjMax during gaming: these usually point to a cooling problem.
- GPU under gaming load: about 65–85°C is typical. Many cards are comfortable below about 80°C.
- GPU at or near 90°C on the main sensor: the card is close to its limit, so check its cooling.
A chip that reaches its limit does not die on the spot. It slows down to protect itself. Repeated high readings, though, mean the cooling needs attention.
Safe ranges by part
These are typical ranges, not limits. Check the spec for your exact part.
| Part | Typical idle | Typical under load | Look closer when |
|---|---|---|---|
| CPU | 30–50°C | 60–85°C in games | Sustained readings in the 90s, or within 5–10°C of TjMax in games |
| GPU (edge) | 30–50°C | 65–85°C | Sustained readings near 90°C |
| GPU hotspot | Close to edge | 10–25°C above edge | The gap grows past about 30°C |
| GPU memory (GDDR6X) | Not usually reported | 90–105°C is common | Throttling near 110°C |
| Motherboard / System | 25–45°C | About 50–55°C | Readings that keep climbing under light load |
| VRM | Not usually a concern | 60–90°C under load | Sustained readings above about 100°C |
| NVMe SSD | Well below 70°C | Keep under about 70°C | Throttling around 70–80°C, varies by drive |
| Hard drive | 25–45°C | Keep below about 50–55°C | Sustained readings above about 50–55°C shorten its life |
What TjMax means for “too hot”
TjMax is the throttle point, not the damage point. When a core reaches it, the chip lowers its clock speed and voltage. Shutdown happens far above that point, so a CPU at its limit is usually slowing down, not failing. See what TjMax means for values by family.
Signs a PC is running too hot
- Fans louder than usual, or running at high speed for long periods
- Lower frame rates or sluggish performance after a few minutes of play, which is typical of thermal throttling
- Random crashes, freezes or blue screens with CPU-related errors
- Sudden shutdowns or restarts under load
- A CPU over-temperature warning at startup
If you see these, confirm the cause with thermal throttling checks before you change anything.
Readings that look alarming but are normal
- Ryzen at 95°C under an all-core stress test. This is the chip’s designed boost limit for many Ryzen 7000 and 9000 models.
- Laptops at 90–100°C under load. Many laptop CPUs are designed to run this hot.
- Idle spikes. Some Ryzen chips briefly spike by 10–15°C at idle as boost kicks in.
- Stress-test numbers. Prime95 with AVX runs hotter than games, so do not judge your gaming cooling by it.
- A warm case. A warm exterior does not mean the internals are overheating.
Also, 80°C alone is not a sign of damage on a modern CPU or GPU.
What to do if your PC runs too hot
- Check whether the chip is throttling. Look at clock speeds under real load, not just the temperature.
- Clean the dust from the heatsink, fans and filters.
- Check the cooler. Confirm it is mounted firmly and the paste is in good condition.
- Improve the airflow and the room temperature.
- Try power limits or undervolting if the chip is still too hot.
Repeated sudden shutdowns under load are a reason to stop and check the cooling before you keep using the PC. For the full list of CPU fixes, see how to stop a CPU from overheating. For Ryzen specifics, see AMD CPU overheating. For graphics cards, see what your GPU temperature should be while gaming.