[mesa-users] OpenMP vs. MPI

[Warrick Ball]

Brian et al.,

I thought I'd chip in on this thread, since I also run MESA grids on 
clusters with larger nodes.  I'd written more but it's been covered in 
other replies, so here's my contribution.

In reality, you may wish to compare the performance to see just how much 
slower the cluster cores are.  In my case, my desktop is only about 10-30% 
faster, but I have 8 cores (4 physical cores, hyperthreaded), compared to 
being able to muster many hundreds on the cluster when necessary.

To get some idea of the cores' rough speeds, I use simple arithmetic 
commands like

echo '2^2^23' | time bc > /dev/null
echo "scale=8000; a(1)*4" | time bc -l > /dev/null

which take on the order of a minute.  The last bit is just to suppress the 
actual output.

I'd be interested if anyone in the forum has better command line 
benchmarking tools?


As for core numbers, I usually don't have that many jobs to run, so I use 
four cores.  (With IBM's load-sharing facility, I have to specifically 
request that they're on one core, though.)  This is based on a few 
considerations about the design of our clusters, and also because I use 
ADIPLS, which is serial (right?).  With more cores, eventually I'd be 
doing fast, parallelize steps of stellar evolution, and just waiting for 
the pulsation calculation.

Cheers,
Warrick




On Wed, 24 Sep 2014, Brian Jackson wrote:

> Dear MESA users,
> I'm planning to run a grid of MESA calculations, and I was hoping to
> use the cluster at my home institution to speed things up. However,
> after getting MESA installed on the cluster, the IT folks here sent me
> the following message (copy-pasted since I don't really understand the
> issues involved):
>
> "Spent some more time looking at how MESA functioned and it seems that
> it utilizes the OpenMP library, rather than an MPI library. The issue
> with this is programs that utilize OpenMP run on a shared memory
> model, which means your MESA jobs would be limited to one node (16
> cores). Because our cluster is focused on quantity of processors, the
> speed of each individual processor is most likely not as high as those
> on your workstation over there (2.0 GHz). This means you would likely
> experience better results running MESA locally (and potentially
> investing in a more powerful workstation if need be)."
>
> So it sounds like running MESA on my iMac with a 3.4 GHz processor
> might be faster than on the cluster. Any suggestions/work-arounds?
>
> Thanks,
> Brian
>
> ---------------------------------------------------
> Brian Jackson, Assistant Professor
> Dept. of Physics
> Boise State University
> 1910 University Drive
> Boise ID 83725-1570
>
> E-mail: bjackson at boisestate.edu
> URL: http://www.astrojack.com
> Phone: (208) 426-3723
> ---------------------------------------------------
>
>
>


------------
Warrick Ball
Postdoc, Institut für Astrophysik Göttingen
wball at astro.physik.uni-goettingen.de
+49 (0) 551 39 5069