[molpro-user] Large differences between CPU TIME and REAL TIME

[Kirk Peterson]

Greg,

usually large wall to cpu time ratios can be directly blamed on your I/O system.  What kind of disks and how many do you have on each
compute node and are they setup with raid0 (assuming you have more than 1) ?

regards,

Kirk

On Apr 19, 2011, at 2:26 PM, Gregory Magoon wrote:

> Hello,
> One of our users has noticed large differences between CPU TIME and REAL TIME
> in several runs and I was wondering if anyone had any tips for getting the REAL
> TIME more in line with the CPU TIME.
> One of the more obvious examples of a large time gap is for an mppx frequency
> run on a 48 processor compute node (using all 48 processors):
> PROGRAMS   *        TOTAL      FREQ      OPTG   CCSD(T)        HF       INT
> CPU TIMES  *     20130.37  10666.84   8936.71    501.37     12.08     13.09
> REAL TIME  *    193915.08 SEC
> The real time is over 9 times longer than the CPU time. The full output file for
> this case is attached.
> 
> For comparison, here are some times running on an 8 processor node, with mpp and
> with mppx, respectively.
> 
> PROGRAMS   *        TOTAL      FREQ      OPTG   CCSD(T)        HF       INT
> CPU TIMES  *     83607.95  68025.70  15492.38     85.26      1.28      3.19
> REAL TIME  *    131591.18 SEC
> 
> PROGRAMS   *        TOTAL      FREQ      OPTG   CCSD(T)        HF       INT
> CPU TIMES  *     78444.88  60833.54  16963.24    627.56     10.09     10.34
> REAL TIME  *     84459.03 SEC
> 
> As you can see, in the 8 processor mppx case (which uses the same MOLPRO
> compilation), the real time is much closer to the CPU time and the real time is
> significantly lower than the case with 48 processors.
> 
> I should note that the tuning was done on the 8 processor node, which has a
> different architecture (Intel, vs. the AMD on the 48 proc. nodes) so I'm
> wondering if that is a factor. Could tuning/architecture explain this (fairly
> large) difference?
> 
> I'm happy to provide additional information (e.g. compilation options) that may
> help diagnose the issue.
> 
> Thanks in advance for any tips/insight you can provide,
> Greg<c2h5oh_3.txt>_______________________________________________
> Molpro-user mailing list
> Molpro-user at molpro.net
> http://www.molpro.net/mailman/listinfo/molpro-user