Benchmarks

We use Par4All to parallelize programs for customers and collaborative research projects but we have also used Par4All on some benchmarks. Here are a few results that we are allowed to communicate.

A video demonstration of Par4All:

Stars-PM

Stars-PM is a Particle-Mesh N-body cosmological simulation in C code from the Observatoire Astronomique de Strasbourg and uses FFT in 3D among other things.

Static Compilation Analysis for Host-Accelerator Communication Optimization.

Our results on 20 benchmarks from Polybench suite, 3 from Rodinia, and the application Stars-PM . Measurements were performed on a machine with two Xeon Westmere X5670 (12 cores at 2.93 GHz) and a nvidia gpu Tesla C2050.

Hyantes:

Hyantes is a library to compute neighbourhood population potential with scale control. It is developed by the Mescal team from the Laboratoire Informatique de Grenoble (France), as a part of Hypercarte project.

Hologram simulation

Fresnel is a C program from the Holotetrix company that simulates Fresnel diffraction  in holograms before manufacturing.

SPEC CPU2006 410.bwaves :

bwaves is a computational fluid dynamics Fortran program that simulates blast waves in three dimensional transonic transient laminar viscous flow.

More information on the program itself on http://www.spec.org/cpu2006/Docs/410.bwaves.html

On a  Wild Node, we get a speed-up of 4.5 with a 2 Intel Xeon X5670 @ 2.93GHz (12 cores).

Matrix multiplication:

The classical Hello World in Fortran  can be found in examples/F77_matmul_OpenMP directory of par4all or look in the git here so you can try by yourself.

On a  Wild Node, we get a speed-up of 12.1 (thanks to cache effects) with a 2 Intel Xeon X5670 @ 2.93GHz (12 cores).