An Educational Grid of a MilLion Processors? “Cool!”

EGEE is a big grid, with 90,000 processor cores available to European scientists. DEISA is even bigger, with 120,000 processor cores.
Impressive science can be done on these grid e-Infrastructures. But what if we had a grid of a million processors? Or even of tens of millions?

People power

A dream? No. Actually, there are millions of computers out there, just waiting to become part of the European science e-Infrastructure. Take an average European town, like Krakow in Poland, or Nice in France.
In each of these urban areas, there are about a million PCs at home and in small and medium businesses. And the good news? People are willing to donate unused computing time to science. This was one of the results of a survey conducted by the e-Infrastructures FP7 project EDGeS. This result highlights an opportunity to further strengthen computational e-Infrastructures in Europe.
The chance to run an application on an even larger grid might make many scientists excited. But if that means learning to compose and submit jobs in a new way, their happiness might dwindle. Luckily, the EDGeS project solves this challenge too. EDGeS aims to build a bridge between EGEE-type grids and BOINC or XtremWeb-based “desktop grids”, allowing seamless exchange of jobs and job results.
So what kind of applications are people willing to run on their PCs? According to the EDGeS survey, three scientific fields clearly stand out. The first is medical applications, an area that affects us all. The same is true for the second category: environmental applications. The third application area, also scoring very highly, was educational applications.

Bridging the gap

Educational applications on a grid e-Infrastructure? Perhaps not something that immediately comes to mind. Yet grid-powered education is a “perfect match”. Let us illustrate this with an example: Engineers must learn how to build a bridge. Today’s bridges often have to look like an art object. That’s fine, but it’s difficult to immediately see whether such a construction might also be stable. You need to run complex programs to do the stress analysis. By providing grid-enabled training software for engineers, engineering
students can design a bridge and then press a button that will start calculating the success of their design.
Applications that combine two of the aforementioned popular grid application areas are also possible. For instance, surgical training can be carried out in a simulated virtual reality environment, where medical students can plan and practice using grid-powered calculations.

From the classroom to the future

Kids certainly would find it “cool” to have access to a grid of a million processors during their math, physics or gaming classes. One thing is sure: by the time they reach university, they won’t find it cool anymore; grids will be an ordinary, but integral, part of their educational science e-Infrastructure.