Quickly inspect current CPU frequencies, governors and scaling behavior from the command line. 16.11.2025 | reading time: 3 min Want to know whether the kernel is scaling your CPU up or down? This short guide shows how to read current frequencies and governors from sysfs and with common utilities so you can diagnose performance or battery issues immediately. Quick live check using sysfs and cpupower ```bash # Read one CPU directly from sysfs cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_cur_freq cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor # Per-CPU summary (works on many distros) for i in /sys/devices/system/cpu/cpu[0-9]*; do echo "$i:" $(cat $i/cpufreq/scaling_cur_freq 2>/dev/null || echo N/A) $(cat $i/cpufreq/scaling_governor 2>/dev/null || echo N/A); done # Or with cpupower if installed sudo cpupower frequency-info # Sample output (sysfs summary) #/sys/devices/system/cpu/cpu0: 3400000 performance #/sys/devices/system/cpu/cpu1: 1200000 powersave ``` Read and interpret the values Look at three things: the current frequency, the active governor and the allowed min/max; files named scaling_cur_freq, scaling_governor and scaling_{min,max}_freq reveal this; a "performance" governor usually pins frequency high while "powersave" keeps it low, and a driver like intel_pstate can change the semantics (it may advertise different controls than acpi-cpufreq). Practical checks and uses Use these checks to answer concrete questions: is the system throttling under heat, does battery mode force lower clocks, are some CPUs stuck at a single frequency, or is a bogus governor configured after a kernel update; run the sysfs checks while stressing the CPU to see scaling in action and combine with temperature sensors to correlate thermal throttling. Permissions and virtualization caveats Some files require root to read and tools like cpupower need elevated permissions, and virtual machines or cloud instances may not expose real cpufreq controls so the files report N/A or fixed values; also note that certain drivers operate in "passive" mode where the governor is controlled by a userspace governor such as ondemand or systemd power profiles. Next commands to try If you want to track changes over time, run a short watch or a loop that samples scaling_cur_freq and scaling_governor, or use tools that report residency and frequency histograms to measure how much time the CPU spends at each P-state rather than just instantaneous values. Finish line and further learning You now have the commands to inspect and reason about CPU frequency scaling; explore governors, drivers and automated power profiles next to gain control of performance and efficiency, and consider formal certification like CompTIA Linux+ or LPIC-1 to deepen system knowledge with intensive exam preparation at bitsandbytes.academy. Join Bits & Bytes Academy First class LINUX exam preparation. utilities processes troubleshooting
