Cet article présente un retour d’expérience sur plusieurs années pendant lesquelles nous avons mis en place de la méthodologie agile, principalement Scrum, dans notre contexte de développement et de recherche scientifiques.

In the context of the ADT Simon related to the Inria Project Lab (IPL) C2S@Exa dealing with High Performance Computations (HPC), a lot of efforts were made by the dtk team to develop both API and efficient back-ends for parallel sparse linear algebra. Simultaneously, Tristan Cabel modified the Fortran code TRACES from the Andra Institute in order to reap the benefit of these developments (see this post).

The nef cluster comes with 6 nodes equipped with Nvidia GPUs.

The “Immersive Space steering committee” is sponsoring a 6 month intership to promote the use of immersive space as a demonstration area.

In an undirected graph, a connected component is a subgraph in which any two vertices are connected to each other by paths. Usual algorithms build a graph where each vertex is labeled with its component number. They use a search from an unclassified vertex to build the connected component containing this vertex. The process repeats until all vertices are labeled with a component. Vertices that are not connected form a component with only one vertex. <!–more– In an edge-centric BSP implementation, like GraphX, finding all the connected components is done in parallel using Pregel iterations:

With the advent of the Qt Installer FrameWork (QtIFW), it has never been easier to provide installers for your Qt based applications. Such installers can be either online (i.e. they fetch contents from the internet from repositories during the installation process, thus allowing updates), or offline (i.e. all the contents is bundled into the installer). QtIFW allows installers to be built for different platforms including Windows, MacOS and Linux.