Pin Me

Building Your Own Supercomputer for Bioinformatics

written by: Finn Orfano • edited by: Leigh A. Zaykoski • updated: 6/18/2008

New advances in Grid Engine software and the use of Bioinformatics are increasing the capability to create your own supercomputer.

  • slide 1 of 1

    New and exciting things are happening in the world of Bioinformatics. How would you like to build your own supercomputer using simple grid technology? Not so long ago, it might have seemed nothing but a story of science fiction, but new technology is enabling the ability to create a supercomputer reality. There is a lot of wasted space that can be tapped into with the right know-how and resources.

    What is the grid concept? It is an infrastructure of components such as hardware, software and instruments that can provide dependable and inexpensive access to otherwise unattainable and untapped resources. Three different types of grids exist:

    Cluster – single team, single organization
    Enterprise – multiple teams, single organization
    Global – multiple teams, multiple organizations.

    In order to unleash the potential of grid technology, you need to use the Sun™ ONE Grid Engine host software to distribute jobs and manage resources. The basis of grid engine hosts involves a master host and shadow host. To start, the user submits their hosts to the master host, which produces execution hosts to generate jobs. Queues are developed as a holding container for jobs that need to be executed and also sort the jobs based on a set of pre-determined criteria. Users can be created and added after the execution hosts are established.

    The possibilities of what grid engine technology can do are very impressive. Some benefits include:

    Getting started quickly and cheaply – it doesn’t take a lot to get started, and for a small start-up fee you can be on your way.

    Growing your grid as needed  - you have the capability to add users and the possibility to grow your grid as you see fit.