The Question is:
Do you have any technical documents that supports the argument that a process
can use up to 100% of the CPU if it is available and not in conflict with
other scheduled processes?
The Answer is :
The null job (the idle loop) effectively does this on all OpenVMS
systems. Many OpenVMS systems running the SETI software package
are also configured to operate this way.
The following DCL works quite nicely as an empirical test:
$ COPY SYS$INPUT: SYS$SYSDEVICE:X.MAR
$ MACRO SYS$SYSDEVICE:X.MAR/OBJECT=SYS$SYSDEVICE:X.OBJ
$ LINK SYS$SYSDEVICE:X.OBJ/EXEC=SYS$SYSDEVICE:X.EXE
$ RUN/DETACH/PRIORITY=0/PROCESS_NAME="Null Wannabe" SYS$SYSDEVICE:X