Re: [Gems-users] Why are ocean finish times coming in negative?

[Nilay Vaish <nilay@xxxxxxxxxxx>]

The amount of time taken for GEMS to execute the Ocean benchmark also 
depends upon the number of time steps for which the Ocean simulation is 
carried out.
--
Nilay

On Mon, 12 Apr 2010, lopamudra chatterjee wrote:

> Hi,
>
> I am running Ocean benchmark with 32 processors.
> But the initialization finish time & overall finish time comes in negative.
> But the correct.out file of model output has positive values althigh it is run with 1 processor
>
> Is a negative value wrong?
>
> Another question
> The run is  taking a long time .
> With 16 core , 4 L2 banks ,1 memory and 18X18 grid points it takes about 2 hours whereas I saw some one saying in the newsgroup that it runs for 18 minutes for them with 4 threads on
>
>
>
>
>
>
> 	258 X 258 grid of integers in a 16 core machine.
> I am running on a 1.3 ghz intel core2 duo machine .
> So anyone has any idea why this is taking so long?
>
> Thanks,
> Lopamudra