|
sir, This a
small description on the code that I want to used in my
calculation:
“Starting from
ADF2007.01, the Windows version
of ADF comes bundled with the HP-MPI runtime environment. This means that you
can run parallel ADF jobs on your brand new multi-core Windows machine. By
default, if the NSCM environment variable is not defined, ADF will use all
logical processors present in the system. To avoid this, you need to set NSCM
to 1. You can do this permanently by adding NSCM with the value 1 to the list
of environment variables or on the per-queue basis putting "export
NSCM=1" (without quotes) in the prolog field of the corresponding queue in
ADFjobs. Moreover, you can define two local queues in ADFjobs: one for serial
and one for parallel jobs. The queues may be called, for example, serial and
parallel. You would then need to change the prolog value for the serial queue
only”
--- En date de : Jeu 31.7.08, Herbert Fruchtl
herbert.fruchtl^st-andrews.ac.uk <owner-chemistry],[ccl.net> a
écrit :
De: Herbert Fruchtl
herbert.fruchtl^st-andrews.ac.uk <owner-chemistry],[ccl.net>
Objet:
CCL:G: Duel cords processors and parallel calculations
À: "Abdelghani,
May " <may01dz],[yahoo.fr>
Date: Jeudi 31 Juillet 2008,
11h54
Sent to CCL by: Herbert Fruchtl
[herbert.fruchtl#,#st-andrews.ac.uk]
It would help to know WHAT you want to
do in parallel. To use several
cores at the same time, you need to
1)
Write the program for parallel execution
2) Compile the program for parallel
execution
3) Run it in parallel.
If you have a parallel executable,
i.e. 1) and 2) are taken care of,
this
leaves you with 3), which depends on the parallelization method
(and the
operating system).
- If it's Gaussian, you need %NProcShared=2 in the
input.
- If it's another OpenMP program, you have to set the environment
variable OMP_NUM_THREADS (it depends on the OS how you do that).
- If
it's an MPI program, you need to run it via a parallel startup
program (in
most implementations this is called mpirun, but the syntax
may vary.
RTFM!).
- If it's yet another parallelization method (PVM, GA, ...), RTFM!
But
since anybody who can't be bothered to mention his program is using
Gaussian anyway...
Herbert
may abdelghani
may01dz*|*yahoo.fr wrote:
> dear CCLers,
> In any calculations, a
duel core processors, using only one core for
> this calculation, whereas
the second rest inactive. I ask if it is any
> possibility to help the
tow cores of my processor to sharing a one
>
calculation, like what we do in the parallel calculations (both cores do
> a one task in the same time).
> Thanks
>
>
>
>
>
>
>
>
>
------------------------------------------------------------------------
>
Envoyé avec Yahoo! Mail
>
<http://us.rd.yahoo.com/mailuk/taglines/isp/control/*http://us.rd.yahoo.com/evt=52423/*http://fr.docs.yahoo.com/mail/overview/index.html>;.
>
Une boite mail plus intelligente.
--
Herbert Fruchtl
EaStCHEM
Fellow
School of Chemistry
University of St Andrews
--
The
University of St Andrews is a charity registered in Scotland:
No
SC013532
-= This is automatically added to each message by the
mailing script =-
To recover the email address of the author of the message,
please change
the strange characters on the top line to the ],[ sign. You can
also
look up the X-Original-> From:
line in the mail header.
E-mail to subscribers: CHEMISTRY],[ccl.net or
use:
E-mail to administrators: CHEMISTRY-REQUEST],[ccl.net or use
http://www.ccl.net/cgi-bin/ccl/send_ccl_message
=
Before posting, check wait time at: http://www.ccl.net
Conferences: http://server.ccl.net/chemistry/announcements/conferences/
Search
Messages: http://www.ccl.net/htdig (login: ccl, Password:
search)
http://www.ccl.net/spammers.txt
RTFI: http://www.ccl.net/chemistry/aboutccl/instructions/ |