[Omp] OpenMP and Intel compiler

[Laksono Adhianto]

Hi,

Thanks for your fast reply (it's 11pm CT, and still lot of people work :-)

Here is the result of SWEEP3D, 50x50x50 grid with 2 MPI processes (for pure MPI)
and 2 MPI processes * 2 OpenMP threads (for the hybrid) without optimization
flags:

----- without debugging mode:
MPI		1.80069398880005
OpenMP/MPI	10.8658130168915
----- with debugging mode:
MPI		88.6593250036240
OpenMP/MPI	58.1125251054764

I run this SWEEP3D twice and the result is the approximately the same. However,
this is just preliminary test, perhaps it's too early to draw any conclusion. I
am sure we cannot generalize that using debugging mode in Intel compiler, hybrid
MPI/OpenMP runs faster than pure MPI :-)

I just wonder if other people has the same experience....

Regards,

------
Laksono Adhianto


On Wed, 17 Dec 2003, Shah, Sanjiv wrote:

> Hello,
>
> This is indeed strange, but I am not sure I fully understand the
> situation.
> Could you please clarify whether both the Hybrid MPI/OpenMP program and
> the
> pure MPI program were compiled with debug mode, or only one of them.
>
> It would be clearest if you included a table as follows:
>
> Code		Time
> -------------------
> Hybrid -g	xx
> MPI -g	xx * 2
> Hybrid	yy
> MPI		yy / 3
>
> Is this what you data indicates?  What is the relationship of xx and yy?
>
> Thanks,
> Sanjiv
>
> -----Original Message-----
> From: Omp-bounces at openmp.org [mailto:Omp-bounces at openmp.org] On Behalf
> Of Laksono Adhianto
> Sent: Wednesday, December 17, 2003 10:12 PM
> To: omp at openmp.org
> Subject: [Omp] OpenMP and Intel compiler
>
>
> Hi
>
> I use Intel compiler v7.1 to compile a hybrid MPI/OpenMP and a pure MPI
> program on Linux Intel Itanium2 clusters.
>
> I found out that if the program is compiled in debugging mode (with -g
> option), then the hybrid MPI/OpenMP is twice faster than pure MPI.
> However, if compiled without debugging mode, then pure MPI is three
> times faster then hybrid MPI/OpenMP.
>
> I also tried with SWEEP3D benchmark, and the result is approximately the
> same (i.e. with debugging mode hybrid MPI/OpenMP is much faster than
> pure MPI).
>
> Is this normal or it's just my 'luck' ?
> Well, "intuitively" either with or without debugging mode, the result
> should be the same ?
>
> Regards,
>
> ------
> Laksono Adhianto
>
>
>
> _______________________________________________
> Omp mailing list
> Omp at openmp.org
> http://openmp.org/mailman/listinfo/omp_openmp.org
>
>