0 - 3 of 3 tags for acctg

Hi all,
We're looking at reporting on CPU usage and trying to understand the difference between the CPU metrics in each table. We have PDCR to archive the acctg table in hourly and daily blocks.

When comparing CPU consumption time between DBQLogTbl.AMPCPUTime and DBC.Acctg against User/Hour/Day, we notice difference in CPU time recorded. Sometime a net difference of 1-2% recorded high or less in Acctg when compared to DBQL.
What makes this difference, is this difference in CPU is respect to Query type or system internal resource.

Is there any way to measure the total AMPCPUTIME consumed by any MLOAD/FLOAD/FEXP job post completion? The values from the dbql(sum of ampcputime for a particular LSN number) does not seem right.
I also tried dbc.acctg. There too, the values seem low.