[FLASH-USERS] Request for reference on performance and scalability

[Nathan Goldbaum]

If this is for an application for compute time, I'd strongly suggest
running the test on the hardware you intend to run on and using the problem
you would like to run in production (to as much of an extent that is
possible). Most supercomputing centers have a way to request discretionary
allocations for these sorts of scaling studies.

On Thu, Oct 11, 2018 at 12:24 PM Carlo Graziani <carlo at oddjob.uchicago.edu>
wrote:

> There's
>
> @inproceedings{cordery2013analysis,
>     title={Analysis of Cray XC30 performance using Trinity-NERSC-8
> benchmarks and comparison with Cray XE6 and IBM BG/Q},
>     author={Cordery, Matthew J and Austin, Brian and Wassermann, HJ and
> Daley, Christopher S and Wright, Nicholas J and Hammond, Simon D and
> Doerfler, Douglas},
>     booktitle={International Workshop on Performance Modeling,
> Benchmarking
> and Simulation of High Performance Computer Systems},
>     pages={52--72},
>     year={2013},
>     organization={Springer}
> }
>
> which is BiBTeX for
> https://www.dcs.warwick.ac.uk/~sdh/pmbs13papers/pmbs0900.pdf.
>
> There's also the following very detailed presentation on getting Flash
> ready for IBM BG/Q from 2013:
>
> https://www.alcf.anl.gov/files/cdaley_miracon_talk_0.pdf
>
> Carlo
>
> On 10/11/18 10:03 AM, Marissa Adams wrote:
> > Hi Ernesto,
> >
> > I don't know of any specific publications off the bat that are straight
> up
> > on the topic of MPI scalability/performance using FLASH. Seems most
> people
> > may discuss these things strategically in a science publication so that
> > work gets cited too -- so you may want to look around at some of the
> > earlier publications in the 2000s
> > (http://flash.uchicago.edu/site/publications/flash_pubs.shtml).
> >
> > Albeit section 8 of the Fryxell et al. (2000) paper may be what you're
> > looking for, but I am not sure if it necessarily pertains to current
> > versions of the code/their performance. Hopefully, Klaus or someone else
> at
> > the Center, or someone else on the user list, knows more -- but this is
> > what I know of.
> >
> > http://iopscience.iop.org/article/10.1086/317361
> >
> > Best of luck,
> > Marissa
> >
> > ---
> > Marissa Adams
> > PhD Student
> > Website: http://www.pas.rochester.edu/~madams
> > <http://www.pas.rochester.edu/~madams/>
> > University of Rochester
> > Department of Physics & Astronomy
> > 371 Bausch & Lomb Hall
> >
> >
> > On Thu, Oct 11, 2018 at 10:53 AM Ernesto Zurbriggen <
> ezurbriggen at unc.edu.ar
> > <mailto:ezurbriggen at unc.edu.ar>> wrote:
> >
> >     Hi all!
> >     Does anybody know if there is a (preferently refereed) reference that
> >     measures the performance and MPI scalability of the Flash code in PCs
> >     and clusters?
> >     I have been looking for it but I got nothing. The reference is needed
> >     for a project.
> >
> >     Thank you!
> >
> >     --
> >     /Dr. Ernesto Zurbriggen/
> >     /
> >     /
> >     /Consejo Nacional de Investigaciones Científicas y Técnicas
> >     //(CONICET)//, Instituto de Astronomía Teórica y Experimental
> (IATE)./
> >     /
> >     /
> >     /Universidad Nacional de Córdoba //(UNC)//, //Observatorio
> Astronómico
> >     de Córdoba //(OAC)//./
> >     /
> >     /
> >     /Teléfono: +54 0351 4331064-5, interno 222./
> >     /_Laprida 854, Bº Observatorio (CP 5000), Córdoba, Argentina_.
> >     /
> >
>
>
> --
> Carlo Graziani                                 (630) 252-1543 (Voice)
> Argonne National Laboratory -------------------------------------
> carlo at mcs.anl.gov           | The less a statesman amounts to, the
>                               | more he loves the flag. -- Kin Hubbard
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20181011/51ff2138/attachment-0001.htm>