Replies: 1 comment 6 replies
-
By default, the CPU and GPU scaling governor is ondemand and causes significant performance issues. I have a systemd service that will set both to performance to counter this issue. I don't know much about the kernel, but it makes sense due to the previously mentioned scaling governor issues. Do you know if this has any meaningful performance impact? |
Beta Was this translation helpful? Give feedback.
6 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Should EAS (Energy Aware Scheduling) be disabled? EAS is a feature implemented in the core Linux kernel scheduler framework to support energy-aware task placement (e.g when to place tasks on a big or little CPU core to save energy). However, it adds a significant amount of extra latency due to expensive calculation for each task placement decision
Since this board is powered using AC power, my suggestion is to simply disable the feature by simply echoing 0 to
/proc/sys/kernel/sched_energy_aware
to switch to faster CFS (Completely Fair Scheduler) scheduling pathBeta Was this translation helpful? Give feedback.
All reactions