Re: Disturbing wide variation in execution time

[Eric Piel <Eric.Piel@lifl.fr>]

(please do not top-post)

07.07.2005 10:03, Sheo Shanker Prasad wrote/a écrit:
> Dear David,
> 
> The program is an atmospheric chemistry-transport modeling code that computes 
> the distributions of atmospheric species (e.g., ozone) as a function of 
> latitude and altitude and how that changes with time.
Well, to let us examine your  problem, you need to describe what 
*technicaly* does your program. ie: lots of disk access, lots of 
networks acces, lots of CPU access, allocate huge memory... in 
particular, is it CPU bound or I/O bound? You should also mention how 
tasks/thread is your program composed of.

It's good to describe your hardware: cluster of NUMA computers or just 
one x86? How much memory, what kind of network, how many hard disks, 
which exact version of the kernel....

Cherry on the top, you could include a *small* program which exhibit the 
same problem.

> 
> Thanks for taking time to think about my problem. I greatly appreciate it.
> 
To answer briefly your question, _in general_ a program which takes more 
than a few seconds to execute should take roughly the same time to 
execute all the times.


Eric