[Carpet] Performance measurements

[Erik Schnetter]

On Sep 18, 2006, at 13:15:29, Ian Hinder wrote:

> Are the timing measurements (i.e. grid point updates per processors
> per second) and the physical time per hour calculated since the start
> of the run, or based on the last few iterations? I would suggest that
> the latter would be better, to give an instantaneous measurement of
> the speed that did not depend on the history of the run.

They are accumulated from the start of the run.  In principle, you  
can take two measurements at different times and use their difference  
to calculate an instantaneous measure.

If you want an instantaneous measure, what time scale should be  
used?  Do you want an additional parameter "restart every n  
iterations", or do you want some kind of decaying average?

-erik

-- 
Erik Schnetter <schnetter at cct.lsu.edu>

My email is as private as my paper mail.  I therefore support encrypting
and signing email messages.  Get my PGP key from www.keyserver.net.



-------------- next part --------------
A non-text attachment was scrubbed...
Name: PGP.sig
Type: application/pgp-signature
Size: 186 bytes
Desc: This is a digitally signed message part
Url : /archives/developers/attachments/20060918/45a40198/attachment.pgp