Dear Miguel!
MEEP or FDTD in general would indeed benefit much from GPU's.
I tried it once. There was a freeware solution fastFDTD using nVidia GPU's.
The speed was incredible high. You gain a factor of 10 in speed very easily.
But the software was completely unstable, full of bugs and no
Dear Benjamin!
On our cluster MEEP in 3D with complex fields takes about 1.2 GB per 10^6 mesh
points (this values varies, but you can take it as an estimation).
You have 20 x 6 x 13 x 100^3 = 1560 * 10^6 points
So I would estimate your memory requirements to about 2000GB.
If you don't have
Dear Steven, dear other MEEP-users!
I followed the discussion with interest about NFS and HDF5 because we have
similar experience.
But what are the alternatives to NFS?
If running meep-mpi on a cluster, every process needs access to the output
directory. So you need to share the directory to
Dear Nizamov, dear Ben, dear Meep-users!
We are working at Intel quad core xeon processors and I can confirm only little
acceleration when using more than 1 core on the same CPU (20 - 30%).
To my knowledge the limiting factor is not the computational power of the
processors, it's the interface
Dear Neal!
We are working on Intel quad-core Xeon processors and it's the same.
The limiting factor is not the computational power of the processors,
it's the interface to the system memory. As in Intel processors all
cores share the same front side bus, execution times decrease very
little
Dear Steven, dear Meep users!
I want to measure the energy, that is radiated away from a continuous
source with a Gaussian beam shape. So I use the flux-in-box
procedure. This is my .ctl-file:
(set! force-complex-fields? true)
(define-param PreTime 100)
(define-param Angle 45)
6 matches
Mail list logo
