# Efficient wake-up latency with power QoS APIs

The power management quality-of-service (PM QoS) API allows drivers to
specify the latency requirement (in microseconds).

If the wake-up latency of a CPU core sleep state is longer than the specified QoS
latency, then the CPU core doesn’t enter the respective sleep state.

For more information about APIs, see [index:
kernel/git/torvalds/linux.git](https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/power/pm_qos_interface.rst?h=v6.6).

## Next steps

- [socreference-power-measurement](https://docs.qualcomm.com/doc/80-70030-30/topic/socreference-power-measurement.html#socreference-power-measurement)

Last Published: Jun 22, 2026

[Previous Topic
Monitor Suspend Power Management](https://docs.qualcomm.com/bundle/publicresource/80-70030-30/topics/suspend-power-management-state.md) [Next Topic
Reference power measurement for system-on-chip](https://docs.qualcomm.com/bundle/publicresource/80-70030-30/topics/socreference-power-measurement.md)