[cam-users] Changes to DTime

[Peter Paul Smolka]

On Fri, 17 Oct 2003 bhuvana@metsun1.met.sjsu.edu wrote:

> Date: Fri, 17 Oct 2003 14:09:17 -0700 (PDT)
> From: bhuvana@metsun1.met.sjsu.edu
> To: cam-users@ucar.edu
> Subject: [cam-users] Changes to DTime
>
> Hi cam users,
> I recently performed a 1 day default(1950) run using CAM2.0 in a single
> processor Linux pc with Portland
> compilers: Eulerian dynamics,64x128,SMP and SPMD both disabled.
> The run was successful , output file looks fine. But the one day run took
> me about 15 minutes. This seems unacceptable as far as my needs go because
> I wish to have  a 50 year run.I know that this might be a processor issue.

Dear Bhuvana,

this is nice to hear as the ccm3.6 needs on a PIII/500 about 18
minutes a day (128x64, T42, dom).

If you think about the processor: Fast Bus (533 MHz or better),
2.6 GHz or so and fast RAM (f. ex. RAMBUS) should help you.

For longer times (e.g. 100 ka, potentially 400 ka):

Wait until in geology it is made faster.

Right now it cannot be promised but good chances for above on a PC
exist (T42 first dom later ccsm aimed at)..

> But with my current availability, is there anything I can do to accomodate
> a 50 year run? For example, can I change resolution + timestep, etc? I am
> not sure how this can be done.
> I was wondering if anyone can provide some guidance here.

Comparable to groundwater modeling changing the timestep may or may
not change the results of the implemented equations.

Regarding resolution: When (1986ff) we dealt with T21 resolution
there was from the geological point of view always the desire for
more (e.g. T42). Since T42 it is for most of the questions
sufficiently realistical. Thus I would for production runs not
decrease the resolution.

I would simply run it as it is and if the disk might spill over
burn the interim results on CD/DVD or "beam" them via /dfs to your
computer center (many have a distributed filesystem to be accessed from
anywhere).

Alternatively you might wish to look into the archive of existing
runs at NCAR (= so you can save the present-day control run if your project
has a deadline).

> Regards.
> Bhuvana

Best regards, Peter


>
> _______________________________________________
> cam-users mailing list
> cam-users@ucar.edu
> http://mailman.ucar.edu/mailman/listinfo/cam-users
>

**********************************************************************
Dr. Peter P. Smolka
University Muenster
Geological Institute
Corrensstr. 24
D-48149 Muenster

Tel.: +49/251/833-3989   +49/2533/4401
Fax:  +49/251/833-3989   +49/2533/4401
E-Mail: smolka@uni-muenster.de
E-Mail: PSmolka@T-Online.de
**********************************************************************