pNetCDF and WRF

[Robert Latham]

On Mon, Oct 30, 2006 at 12:18:41PM -0700, John Michalakes wrote:
> Here are some preliminary results from the NCAR IBM Power5 system
> (bluevista) comparing parallel NetCDF vs regular NetCDF output in the WRF
> model. This is for one of the standard WRF benchmark cases (12km CONUS; see
> http://www.mmm.ucar.edu/wrf/WG2/bench) where each write is about 250 MB.  I
> timed writes to a new NetCDF file and writes to an already existing NetCDF
> file separately.

Wow, I'm really surprised to see such a large penalty for creating the
file.  What file system is this?  I see serial netcdf pays a small
penalty, but it's more like 10% instead of PNetCDF's 50% penalty.  

> If anyone's intersted in obtaining this new version of WRF that uses
> pNetCDF for your own testing, please let me know.

I sure am interested.  I might not be able to look at it very much
until some time after supercomputing, but we would definitely like
anohter workload we can use for our pnetcdf tuning efforts.

==rob

-- 
Rob Latham
Mathematics and Computer Science Division    A215 0178 EA2D B059 8CDF
Argonne National Lab, IL USA                 B29D F333 664A 4280 315B