Hi,
I've running some simulations in different computers. However, I've noticed that they are taking longer than they used to take. I mean, running the same cas in two different computers (same computer model) takes about 30% more time with the one that has v7p2 installed. I add that the computers are both -almost- new, so I hope they should run in a very similar time.
My concern is that they might be computing things in a different way or is it that really v7p2 is slower?
I must say that I'm using the same cas with both installations, it was originally created for the v7p0 computer. Since my fudaa version didn't have v7p2 as option, I used the same because I thought there were no important differences among them.
I'd like to point out something that happened to me with this case:
In my v7p0 cas, I used the keyword:
INITIAL VALUES OF TRACERS =0
But at the same time I modified the condin.f to have some special IC of tracers. For some reason I thought that the keyword should be there even if they are defined in the fortran file. My v7p0 installation did the right thing and set the conditions as prescribed in the fortran file, ignoring the cas.
But my v7p2 installation ignored my fortran and set the tracer initial condition as in the cas! As soon as I deleted the INITIAL VALUES OF TRACERS option, it did it as in the fortran.
So from there, I'm now concerned because it takes more time when computing and this difference is not listed in the new things of v7p2.
Any comments are appreciated.
Thanks,
JR
PS: I'm running in windows 10