Re: st: RE: Will Stata/MP speed up running multiple dofiles inbatch mode?

[Neil Shephard <nshephard@gmail.com>]

On Thu, Jan 20, 2011 at 11:44 AM, Alex Eapen <alex.eapen@sydney.edu.au> wrote:
> Thanks for the further explanation and links, Neil. I intended to say in my earlier reply that the commands in each of my do-files are -summarize- and -regress-. These commands, according to the Stata/MP performance report, run faster on multiple cores. So my individual do-files will indeed run faster in Stata/MP and, as you say, the entire exercise should be faster compared to Stata/SE. I understand your point that what might make batch jobs run faster in general is better hardware (i.e., more memory or a faster CPU which makes available more resources to the concurrent do-files) than an upgrade to Stata/MP.

Ok, that clears up that there is no real misunderstanding.

Personally I couldn't justify the outlay for such simple commands.
Why not just get your scripts set up and then leave them to run
sequentially overnight?

That way each script will complete in the minimum amount of time
because its not vying for RAM/CPU with other instances of Stata.

If it were a scenario like bootstrapping where instead of running all
1000 bootstraps on the same processor you would essentially have
Stata/MP sending off 250 to each processor then yes it would be a
benefit.

Neil





-- 
“Truth in science can be defined as the working hypothesis best suited
to open the way to the next better one.” - Konrad Lorenz

Email - nshephard@gmail.com
Website - http://kimura.no-ip.org/
Photos - http://www.flickr.com/photos/slackline/

*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/