Grid'5000 user report for

Jump to: navigation, search

User information

( user)
More user information in the user management interface.


  • YML Workflow environment performance evaluation (Programming) [in progress]

    The context of this experiments is detailed here.

    The experiments describe here makes use of two test applications:

    • a distributed sort which sorts a huge amount of data in blocks. The dataset consists in blocks of values to be sorted. Each block contains an arbitrary number of values. The dataset can be configured for each experiments by defining the number of blocks and the number of value in each block. Using this simple application we are able to create arbitrary large dataset to stress YML on several aspect such as the amount of potential parallel tasks in the application as well as the ratio between the computations time and the communication time.
    • an image manipulation application with a more complex dependency pattern. This application apply several transformation stage to an image split in square blocks. The filtering pass needs to deal with the blocks composing the neighborhood of an image at the previous stage. This application is a typical illustration of the added value of YML over XtremWeb which is provided with a povray parallel rendering demonstration. The image manipulation application can be configured in several ways: We have the ability to adjust our experiments on the number of blocks composing an image as well as the size of each block. We can also configure the number of transformation stage and for each stage the transformation operation to use. At the time being we implemented a single image filtering tool, a Gaussian blur filtering algorithm.


    The first goal of this experiments series is to check the scalability of YML and to measure several aspect related to the usage of YML. Measurements includes :

    • the overhead introduce by YML using the various middleware supported,
    • the overhead due to the management of a middleware including a great number of peer.
    • the effect of scheduling several YML application on the same middleware at the same time.

    The second goal of this experiments series is to evaluate the behaviour of YML in data intensive application. For this goal we will configure our applications so that the amount of time spent in computation is small in regards to the time spent on communication.

    This experiments is underway at Orsay site. A multisite evaluation is also planed at a later time. Results are not yet available and will be added soon.
    Results: not yet available
    More information here
  • High performance numerical computing with YML (Application) [planned]

    The context of this experiments is detailed here.

    Numerous problems can be modeled using linear algebra application such as Linear system solving and Eigenvalue problems. In order to implements solver for large problem, the CNI team of PRiSM laboratory has evaluated various approach to implements hybrid method such as Multiply Explicitly Restarted Arnoldi Method(MERAM). This approach includes : Sequential and Parallel (MPI) object oriented library on cluster and parallel HPC computers, Problem Solving Environment using Netsolve and more recently workflow application description using YML.

    An iterative method computes its result by creating a succession of partial results. Those results are approximation of the solution of the problem. Each iteration try to produce a new solution based on the results acquired at the previous stage. The performance of an iterative method is measured based on the number of iteration needed to reach the solution. A way to improve the performance is to combined several numerical method together in order to decrease the number of iteration needed to converge. MERAM is such a method and use multiple instance of the ERAM process which exchange intermediate results asynchronously. With the previous approaches we have been using based on MPI we have not been able to implement scalable version of MERAM and to use them with a high number of collaborating ERAM process. The use of YML as our main application description allowed us to easily change the number of ERAM processes collaborating.

    The aim of these experiments is to evaluate the effect of a large number of ERAM process on the number of iteration needed to converge. We also want to evaluate our methods on huge problem with sparse matrix of order 10^6 and more. Results of this experiments will hopefully gives us more knowledge on the scalability of hybrid methods. We will also be able to evaluate YML with complex graph and realistic applications on real problem. For those experiments we will use matrix generator from Matrix Market in order to be able to use arbitrary huge matrices.

    More information here



      Success stories and benefits from Grid'5000

      last update: 2007-02-28 16:48:16

      Personal tools

      Public Portal
      Users Portal
      Admin portal
      Wiki special pages