Re: [Meep-discuss] Slow execution issue (Andrea Armaroli)

Hi Andrea, It seems that your load is well balanced... Can you tell me how your run scales (ie: runtime for 1 proc, then for 2 procs,4 procs, etc...) in absence of I/O (no hdf5 call)? Sorry I cannot help you more than that, but if the scaling is good for 2 procs and bad after, it means that

Re: [Meep-discuss] Slow execution issue (Andrea Armaroli)

Hi Andrea, Can you be a bit more specific about the number of processors you are using on your cluster? From the numbers you gave, it looks like you're using only 1 proc. I have a problem with meep-mpi, which only shows proper scaling with 2 processors, but not more (see ) Can you give

Re: [Meep-discuss] meep-mpi 3d ring transmission problem

Hi Markos, I also have a problem with meep-mpi scaling, described in thefollowing http://www.mail-archive.com/meep-discuss@ab-initio.mit.edu/msg03490.html The first thing you want to do in order to know how good or bad meep-mpi scales on more than 1 processor in remove all I/O. HDF5 mpi see

Re: [Meep-discuss] meep-mpi scaling perf with more than 2 processors

6411088k total, 6047088k used, 10364000k free, 256k buffers Swap:0k total,0k used,0k free, 304812k cached PID USER PR NI VIRT RES SHR S %CPU %MEMTIME+ COMMAND 16811 gdemesy 25 0 1243m 1.1g 6804 R 100.1 7.0 1:19.76 meep-mpi 16812 gdemesy 25

Re: [Meep-discuss] meep-mpi scaling perf with more than 2 processors

4987232k free, 256k buffers Swap:0k total,0k used,0k free, 275088k cached PID USER PR NI VIRT RES SHR S %CPU %MEMTIME+ COMMAND 5907 gdemesy 25 0 946m 751m 6840 R 104.9 4.7 2:06.09 meep-mpi 5909 gdemesy 25 0 949m 755m 7000 R 104.9 4.7

Re: [Meep-discuss] meep-mpi scaling perf with more than 2 processors

Hi Shavkat, thanks a lot for your quick answer. 1)You are right, I didn't mention the layout of the cluster. For now, I'm working on one node of a larger machine based on Intel's Nehalem architecture (I'm trying to evaluate performance on only one node before using more). Each node has 16Gb

[Meep-discuss] meep-mpi scaling perf with more than 2 processors

Dear Meep users and developer, I'm getting strange scaling performance using meep-mpi compiled with IntelMPI on our cluster. When I go from 1 to 2 processors, I'm getting an almost ideal scaling (i.e. runtime is divided by almost 2 as shown below for various problem sizes), but the scaling

[Meep-discuss] numerical material-function and eps-averaging

Dear Meep users and developer, I am eventually willing to model excentric shapes using meep, for which I believe I need to use "material-function". However, it seems that using material-function automatically turns off eps-averaging when using libctl (analytical averaging at least). But I rea

Re: [Meep-discuss] Slab: Comparison with analytical results

Actually, setting resolution to 200 leads to a closer result (cf image link below). There is still a frequency shift of a bit inferior to 0.005 ... But it also means that I am spatially sampling the smallest wavelength propagating in freespace by 200/(1.125) = 180 points !!! I was expecti

[Meep-discuss] Slab: Comparison with analytical results

Dear Meep Users and Developer, I'm new to Meep and I'd like to thank the developer for this nice soft. As a first exercice, I'm trying to retrieve the intensity transmission and reflexion coefficients of a dielectric slab (Fabry-Perrot). I'm dealing with this 1D problem in 2D. The setup is