Comments on: Digging Into The Top500 http://www.linux-mag.com/id/7922/ Open Source, Open Standards Sat, 05 Oct 2013 13:48:18 +0000 hourly 1 http://wordpress.org/?v=3.1 By: scemama http://www.linux-mag.com/id/7922/#comment-8772 scemama Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7922/#comment-8772 <p>Hi,<br /> here is an answer your question about who uses a large number of processors for one job:</p> <p>Stochastic methods are very well adapted to using thousands of processors since an unlimited number tasks can be run independently and asynchronously. In addition, fault-tolerance is easy to achieve: if one task doesn't transmit its output it affects only the error bar on the total result and not the average. These methods are quite frequent (namely in physics, chemistry, finance...) and will probably spread even more in the future due to its massively parallelizable property.</p> <p>For example, this page shows an application of Quantum Monte Carlo<br /> on the EGEE grid : http://qmcchem.ups-tlse.fr/index.php?title=Large-scale_quantum_Monte_Carlo_electronic_structure_calculations_on_the_EGEE_grid<br /> In addition to grids, this code runs routinely on 512 cores (for a single calculation) with a nearly perfect speedup. It could run on much more if we could have access to more cores!</p> <p>Anthony </p> Hi,
here is an answer your question about who uses a large number of processors for one job:

Stochastic methods are very well adapted to using thousands of processors since an unlimited number tasks can be run independently and asynchronously. In addition, fault-tolerance is easy to achieve: if one task doesn’t transmit its output it affects only the error bar on the total result and not the average. These methods are quite frequent (namely in physics, chemistry, finance…) and will probably spread even more in the future due to its massively parallelizable property.

For example, this page shows an application of Quantum Monte Carlo
on the EGEE grid : http://qmcchem.ups-tlse.fr/index.php?title=Large-scale_quantum_Monte_Carlo_electronic_structure_calculations_on_the_EGEE_grid
In addition to grids, this code runs routinely on 512 cores (for a single calculation) with a nearly perfect speedup. It could run on much more if we could have access to more cores!

Anthony

]]> By: schandok http://www.linux-mag.com/id/7922/#comment-8773 schandok Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7922/#comment-8773 <p>Nice summary about on the direction HPC clusters are taking. </p> Nice summary about on the direction HPC clusters are taking.

]]> By: laytonjb http://www.linux-mag.com/id/7922/#comment-8774 laytonjb Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7922/#comment-8774 <p>@scemama,</p> <p>I think Doug is referring to a single application running on 128,000 cores using MPI. </p> <p>Jeff </p> @scemama,

I think Doug is referring to a single application running on 128,000 cores using MPI.

Jeff

]]> By: solanum http://www.linux-mag.com/id/7922/#comment-8775 solanum Wed, 30 Nov -0001 00:00:00 +0000 http://www.linux-mag.com/id/7922/#comment-8775 <p>Boinc uses more then 128K cores. Their cluster just has some latency issues. :) </p> Boinc uses more then 128K cores. Their cluster just has some latency issues. :)

]]>