There seems to have been some (stupid) change to the Linux kernel, where the CPU are stymied to a low power setting.
running cpufreq-info I get: The governor "ondemand" may decide which speed to use
I see that I can change one core like this: sudo cpufreq-set -c95 -g performance
But how do I set ALL cpu cores to PERFORMANCE permanently (persist across reboots)?
In the file:
/etc/default/cpufrequtils Make a setting like this:
GOVERNOR="performance" Then restart the service:
sudo systemctl restart cpufrequtils You can later check your settings with:
cpufreq-info I would use tuned which is a systemd service. It would solve an issues regarding not persisting after reboots.
Below are the profiles that are shown from tuned-2.22.1-4.el8_10.1.noarch installed in RHEL-8.10.
I would do tuned-adm profile latency-performance
tuned-adm list Available profiles: - accelerator-performance - Throughput performance based tuning with disabled higher latency STOP states - atomic-guest - Optimize virtual guests based on the Atomic variant - atomic-host - Optimize bare metal systems running the Atomic variant - aws - Optimize for aws ec2 instances - balanced - General non-specialized tuned profile - cpu-partitioning - Optimize for CPU partitioning - cpu-partitioning-powersave - Optimize for CPU partitioning with additional powersave - default - Legacy default tuned profile - desktop - Optimize for the desktop use-case - desktop-powersave - Optmize for the desktop use-case with power saving - enterprise-storage - Legacy profile for RHEL6, for RHEL7, please use throughput-performance profile - epyc-eda - Optimize for EDA compute workloads on AMD EPYC CPUs - hpc-compute - Optimize for HPC compute workloads - intel-sst - Configure for Intel Speed Select Base Frequency - laptop-ac-powersave - Optimize for laptop with power savings - laptop-battery-powersave - Optimize laptop profile with more aggressive power saving - latency-performance - Optimize for deterministic performance at the cost of increased power consumption - mssql - Optimize for Microsoft SQL Server - network-latency - Optimize for deterministic performance at the cost of increased power consumption, focused on low latency network performance - network-throughput - Optimize for streaming network throughput, generally only necessary on older CPUs or 40G+ networks - optimize-serial-console - Optimize for serial console use. - oracle - Optimize for Oracle RDBMS - postgresql - Optimize for PostgreSQL server - powersave - Optimize for low power consumption - server-powersave - Optimize for server power savings - spindown-disk - Optimize for power saving by spinning-down rotational disks - throughput-performance - Broadly applicable tuning that provides excellent performance across a variety of common server workloads - virtual-guest - Optimize for running inside a virtual guest - virtual-host - Optimize for running KVM guests Current active profile: latency-performance I'm not sure if it'll work on x86 computers, but it does work on the RaspberryPi:
echo performance | sudo tee /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor Please note that this won't persist across reboots. So you'll have to run this again every time (See comment to my answer).
/sys does not persist across reboots. /sys is a live-only filesystem that exposes kernel interfaces. What you read and write there exists only in memory.
ondemandorpowersaveare the correct choice for most users, including CPU-heavy workloads. Setting the governor toperformancewill make your computer a lot hotter with little tangible benefit.