How to Check CPU Temperature on Linux

As an experienced Linux developer and system administrator, monitoring CPU temperature is an essential practice I prioritize for all of my systems. When the CPU reaches excessively high temperatures, it can cause system crashes, lockups, hardware damage and permanent errors. Carefully watching CPU heat levels and being proactive about mitigating overheating has helped me resolve many issues over the years.

In this comprehensive guide, I will explore the ins and outs of checking and monitoring CPU temperature in Linux.

What Causes CPU Overheating?

Before diving into the monitoring methods, let‘s look at why CPUs get hot in the first place and common culprits behind temperature spikes.

Modern CPUs generate significant amounts of heat due to their increasing complexity and high clock speeds. Heat is a byproduct of operating the billions of tiny transistors that make up a CPU. The harder a CPU works, the more power it consumes, and the hotter it gets.

Activities that drive up CPU utilization push chips to work extra hard, like:

• Compiling large programs
• Running virtual machines
• Executing scientific/math calculations
• Rendering graphics and video
• Gaming

Besides heavy usage loads, insufficient cooling and underlying hardware issues also cause overheating:

Inadequate Cooling

Laptops and small form factor PCs often have cramped cases with limited airflow. Insufficient fans or obstructed vents prevent hot CPU air from adequately escaping. Aftermarket cooling pads and stands can help improve circulation around a laptop. For desktops, adding more fans or reconfiguring air flow inside the case helps expel rising heat from the CPU and components.

Faulty Hardware

Defective CPU heatsinks result in improper heat transfer away from the chip die. This is also true when the thermal interface material/grease has dried out. Thermal throttling fails to engage on some CPUs to slow clocks and voltage when approaching excess temperatures. Flaws at the hardware level typically require replacing the affected components to resolve overheating.

Identifying the activities triggering spikes along with ruling cooling issues and hardware faults helps narrow down the source of possible overheating. Now let‘s explore helpful utilities to precisely monitor your CPU temperature in Linux.

Monitoring Utilities for CPU Temperature

Here are reliable tools I routinely utilize to check on CPU heat levels across my Linux computers:

1. sensors

The sensors program leverages lm_sensors to query hardware monitoring chips called sensor chips to report current CPU temperature. Popular sensor chip manufacturers include Winbond, IT87, Adm1024, Via686a, etc. My AMD Ryzen desktop runs an IT8688E chip for example.

To install sensors:

sudo apt update
sudo apt install lm-sensors 
sudo sensors-detect

Running sensors-detect scans hardware and does automatic chip detection to find available sensors, which gets written to /etc/sensors.conf.

With the config set up, run sensors:

sensors

Temperature read outs for CPU cores show next to labels like "Tctl" or "Tdie".

The lm_sensors module also ties into the hwmon or hardware monitoring kernel subsystem. Temperature values are accessible under /sys/class/hwmon.

This example reads the temp of the first processor on my 12-core Ryzen:

cat /sys/class/hwmon/hwmon2/temp1_input 

2. psensor

A favorite graphical app of mine for simple temperature monitoring is psensor. It has an intuitive interface that plots nice graphs over time for visualizing spikes.

psensor installation:

sudo apt update
sudo apt install psensor

Launch psensor to have it auto-detect components and begin plotting temperatures. Customize appearance under preferences like enabling RGB mode or using a different color palette.

Key things I use psensor for include setting temperature thresholds to trigger notifications when exceeding limits and actively monitoring during performance testing. For example, running demanding benchmarks like Phoronix or stress testing with s-tui.

3. htop

No list of Linux process and system monitoring tools is complete without htop! The nifty interactive process viewer packs a surprising amount of metric monitoring features as well.

Install htop using:

sudo apt update
sudo apt install htop

Run htop then hit F2 to access setup options. Check "Sensors" under "Meters" to enable temperature readings. Back on the main display, temperatures now show inline by component name!

I routinely have htop running to glimpse overall system health, keeping an eye on any CPU temperature shifts.

GNOME Shell Extensions

For GNOME desktop users, extensions provide superb fine-grained, constant temperature monitoring right in your top bar. Some top options include:

CPU Temperature Indicator – Simple graphical indicator and numeric display for CPU temps.

CPU Temperature Monitoring – Adds percentage indicators per core and warning color changes based on threshold breaches.

Visit https://extensions.gnome.org to search full catalog and manage extensions. Pay attention to compatibility with your GNOME version before installing.

………………………..

Strategies to Reduce Linux CPU Overheating

Okay, with your enhanced Linux thermal monitoring toolbelt equipped, what can actually be done to curb overheating? Here are proactive optimization and troubleshooting steps I take:

1. Tune CPU Performance Settings

Scale back your CPU to run slower using Linux‘s CPUFreq subsystem and governors. I create a power optimized profile via Tuned lowering maximum processor frequency. Or quickly change governor to ‘powersave‘ forcing lowest possible clock speeds. Both dramatically limit heat production.

2. Improve Air Flow

If dealing with desktop tower heat issues, open up the case and examine fan placement and component spacing. Dust buildup greatly inhibits airflow and heat dissipation. Carefully clean CPU fans and heatsinks so hot air properly exhausts. Reconfigure fans to increase internal case flow circulation.

Consider adding CPU or chassis fans to improve CFM (cubic feet per minute) airflow volumes based on case size. Especially helpful for overclockers pushing CPUs extra hard.

3. Replace Thermal Compound

The thermal interface material/grease between a CPU and heatsink dries out over time. This prevents heat from transferring away from the CPU effectively. Every 2 years, I open up my CPUs and replace the compound for optimal heat conduction. Arctic Silver or Noctua NT-H1 work fantastic for lowering thermals.

4. Address Faulty Hardware

If you‘ve addressed the software-based avenues and excessive CPU temperatures still plague your system, underlying hardware likely needs replacement. Diagnosing precisely which component depends on the symptoms:

• System crashes under load – CPU overheating and unable to throttle down. Replace CPU or heatsink mounting mechanism.
• Sudden extreme spikes – Heatsink separated from CPU surface. Replace heatsink with fresh thermal paste application.
• High idle temperatures – Malfunctioning heatsink unable to dissipate ambient heat. Replace heatsink.

Chip manufacturers like Intel and AMD document maximum safe operating temperatures before hardware damage occurs:

+--------------------+-------------------------+
| Manufacturer       | Max Temp Threshold      |  
+--------------------+-------------------------+
| Intel              | 100°C                   |
| AMD Ryzen 3000    | 95°C                    |   
| AMD Ryzen 5000    | 90°C                    |
+--------------------+-------------------------+

Monitoring utilities give you visibility into your CPU‘s temperature exposure based on workload. By combining proactive cooling tactics, performance tuning, replacing worn hardware, and watching sensors – you can achieve stable equilibrium protecting your CPU as a Linux power user!