Hi,
I am not familiar with db2 internal working , and I have a one serious problem to handle

actually I have been using db2 8.2 on win2003 4 core CPU and it was working fine since my heavy data report started tacking 24hrs to complete , So I bought IBM Server with 8 core and 12 GB of RAM
Now problem is using EE db2 9.7 on RHEL 6 , my report process only use 2 CPU out of 8 core and If I initiate second thread manually only then process spreads to 4 CPU,

So problem is why in linux 9.7 my process is limited to 2 CPU in Db2 EE but use 4 CPU on db 8.2 on windows automatically

(and workgroup edition use only one CPU )
I have already run the INTRA_PARALLEL, but no luck

Have you recycled instance after you alter set INTRA_PARALLEL to Yes?
Maybe you also need to alter the db cfg (DFT_DEGREE) to any to let db2 choose query parallel degree.
You can use db2expln or db2exfmt to see query degree for the specific sql。

what do u mean by recycle , restart ? yes I have

Is DB2 properly licenced? It behaves like an express-C version....
Check with "db2licm -l"

no it is not license yet , but It not express-c it is sure enterprise edition 9.7 I have checked it with db2licm -l

it is under evaluation till we come over this problem

yeah , recycle = restart.
Maybe access plans for your query have been changed when you migrate your system on db2 v9.7。
You can use get snapshot for application to see how much agents associated with your application when runing the data report.
"Maximum associated agents = XX"
and Not all excuting phases of a query can't be processed parallelly ( fetch ,compile,etc)
anyway it will be ealier to dig out what the problem is if you put the exactly access plan here.

The important thing is, how you determine that the DB2 process uses as many CPU cores as you think it does.

As for the differences between DB2 8.2 on Windows and 9.7 on Linux, they have completely different process models and you cannot possibly compare the two.

In the beginning of db2diag.log there is a banner printed that shows how much memory and how many CPU cores the instance actually uses.