Found an article
How is "Compute Processor Queue Length" calculated - CA Knowledge
How about other OS? How is it calculated?
Processor Queue Length means slightly different things on Windows and Unix systems. On Windows, there is a specific counter called Processor Queue Length. On Unix an approximation of this data is the load data, which is what we use to get a comparable value. On Windows the counter is an integer, but the load data in Unix is aggregated over time, and is thus a floating point number. The fact that the processor queue length is in fact load should probably be reflected in the alarms and log, but this has not been implemented at this time.
For AIX, SGI Linux and Solaris, this option monitors system load.
Alarm on Processor Queue Length (For Windows)/Alarm on System Load (For AIX, SGI Linux and Solaris)
Select this alarm setting to check the processor queue length. The processor queue length measures the number of threads that are in the server's processor queue waiting to be executed by the CPU. All servers, whether they have a single CPU, or multiple CPUs, have only one processor queue. The processor queue length is a measurement of the last observed value, and it is not an average of any kind. Alarm messages are generated according to the specified threshold value. Default: 4.
If running on a multi-CPU system, the queued processes will be shared on the number of processors. For example, if running on a system with four processors and using the default Max Queue Length value (4), alarm messages will be generated if the number of queued processes exceeds 16.
To enable the QoS metric QOS_PROC_QUEUE_LEN for each CPU, you are required to add a key system_load_per_cpu with value as Yes under the CPU section through the raw configure option. The probe calculates the system load on Linux, Solaris and AIX as Load/Number of CPU if this key is set to Yes.
Thanks for the elaborate answer.