Understanding Grid Computing: Examples of Grid Computing
written by: Karishma Sundaram•edited by: Rebecca Scudder•updated: 3/29/2010
Grid Computing is still a new concept in computing. Applications requiring huge computational power becoming more prolific, and grids are the only viable solution. Let’s look at some examples of grid computing, and applications where they could make a difference.
slide 1 of 5
A Brief Introduction to Grid Computing
Grid computing is a technology used to harness computing powers from various sources and use them in harmony to achieve a specific goal. The great advantage of grid computing is the ability to significantly reduce the time that is taken to accomplish that goal, thereby increasing efficiency.
Most applications of grid computing are ones where the computing resources of one computing unit prove to be insufficient for the task at hand. The computing unit in question could potentially range from a single personal computer to a supercomputer within a large organization.
For example, a weather forecasting unit would require multiple variables and calculations within the program. Computing various scenarios and determining the probability of each scenario requires a large amount of computing power and time. The data that is required for such a task needs to be current, and the results need to be available within a certain time frame. This is an ideal application for grid computing.
slide 2 of 5
One of the more famous examples of grid computing projects is run by SETI, also known as Search for Extra-Terrestrial Intelligence. The application looks for radio signals or other forms of communication in space, in an effort to prove the existence of extra-terrestrial intelligence.
SETI developed a grid computing middleware where the program could be executed over multiple computers, since the application requires huge computing resources to scan the skies effectively. The infrastructure was designed in such a way that a layman using the Internet could choose to donate their unused computing power to the project. The middleware in known as BOINC (Berkeley Open Infrastructure for Network Computing) and was distributed under a GNU public licence. All a user had to do was download and install the middleware, and their idle processing power was at the disposal of SETI.
BOINC went on to become so popular, it was used for many applications besides SETI@home.
slide 3 of 5
LHC Computing Grid
Another famous application of grid computing technology is the computing grid used to support the Large Hadron Collider at CERN (European Organization for Nuclear Research). The LHC Computing Grid is unlike the SETI@home project, in that it does not use donated processing power. The computing grid is a closed one, since only certain organizations have access to it.
As a prime example of using a grid computing infrastructure to maximize efficiency, the LHC Computing Grid ably demonstrates its efficacy. The collider generates approximately 27 TB of data per day, and all the data has to be scanned for interesting readings – a feat that would be practically impossible even on the faster supercomputer.
slide 4 of 5
NFCR Centre for Computational Drug Discovery
Known as by the moniker ‘Screensaver Lifesaver’, the computational grid at Oxford University’ Centre for Computational Drug Discovery is built along the same principles as the BOINC infrastructure.
In this project, volunteers donate a few of the computing cycles when their screensavers are running. When screensavers are running, the computer is essentially idle, apart from a few background tasks. The project aims to find a cure for cancer using computational methods to screen small molecule structures, otherwise shortening a very lengthy process.
slide 5 of 5
Other Grid Applications
There are many smaller computational grids that exist all over the globe. Typically applications like weather forecasting, protein folding and earthquake simulation are prime candidates for a grid infrastructure. Grids have also been used to render large-scale animation projects, like movies.
There are many uses for computational grids, which are beginning to surface more since the volume of data that is generated, and as a result analyzed at a later stage, grows exponentially every day.