
The IT department atCern
, birthplace of the World Wide Web, is creating the world's
largestcomputing gridto help scientists
study the origins of the universe.
A grid of this type could eventually have
applications in the finance sector. Such a development could
see finacial firms dealing with gigabytes of data per second within
the next five years, analysts said.
The grid, which will go live in June, will pool the processing
power of approximately 100,000 CPUs worldwide. It will process
information at a rate of 1gbps, said Francois Grey, head of Cern's
IT communications team.
"About 20% [of CPUs] will be here at Cern, another 30% will be
in 11 major datacentres, and the rest of the computing power will
come from 250 other centres worldwide," said Grey.
Scientists working on the
Large Hadron Collider project need vast computing power to
process the
mass of data generated by the experiment, which recreates the
"big bang". The experiment will produce roughly 15 petabytes (15
million gigabytes) of data a year - enough to fill 100,000
DVDs.
Grid computing is already being used by Google and Amazon and it
will have applications in the financial sector as processing
demands grow, said analyst firm Gartner.
A key challenge in grid computing is ensuring reliable and
secure access to a widespread IT infrastructure.
Cern co-developed next-generation middleware, called gLite, to
use the processing power of computers reliably in a 24-hour
operation.
"A lot of the academic projects work with existing middleware
products, but when you want to use it 24x7, you discover they have
some bugs," said Grey.
The gLite software authenticates users to the grid and ensures
that contributors of computing resources are not exposed to
security threats.
"What is special about this middleware is it handles all the
issues that arise when you have many organisations sharing
resources. We are talking about legally independent organisations
in many different countries with different laws on how you process
data," said Grey.
Blog: reporting live from Cern >>