Cloud and grid computing yields new business opportunity
The emergence of cloud and grid computing across the globe is giving birth to a new business, one that registered and connects massive networks of computing resources with organizations that require those resources.
In an article published by International Science Grid This Week, author Danielle Venton at EGEE (Enabling Grids for E-Science), an European Commission co-funded project to promote the use of grid computing for academic and business research, says the emergence of applications that require vast computational resources, such as experiments being conducted at CERN’s Large Hadron Collider, has now given birth to so-called ‘service registries’ for computing resources.
These companies, like US-based Digital Ribbon, are now serving as network interconnection points for grid computing resources, linking together different centres with large computational capabilities and offering them up to organizations looking to execute large calculations.
According to Digital Ribbon, it currently has direct access to capacity exceeding 9.5 million CPU hours a month, and indirect access to additional 14 million CPU hours a month. The largest single image system in its constellation of resources has over 4,600 cores, and 2 gigs of bandwidth, Digital Ribbon said, adding that it also offers diverse processors (32bit, 64bit), multiple interconnect technologies (including Infiniband) and one of the largest Windows clusters in the world (1,600+ Cores).
For its part, EGEE is one such source, consisting of 300 sites in 50 countries serving 10,000 users with 80,000 CPU cores around the clock.
DIFFERENT QOS
More importantly, Venton writes that these service registries are not only providing access to computing resources but also serves to categorise the resources according to type and capabilities.
“Computing is moving towards becoming a commodity,” Erik Weaver, Digital Ribbon’s CEO, told Venton. “However it is much more complex than the oft-cited example of electricity.”
According to the article, computing consumers care about more than power, they also care about bandwidth, interconnectivity (how fast individual processing cores can talk to each other) and memory.
It is important to have not just enough resources, but also the right kind. Weaver explained: “If you are making an apple pie, you might want to use Granny Smith, not Fuji apples.”
The EGEE recently conducting a test with Digital Ribbon and successfully ran applications on its network on the resources of Digital Ribbon. The test used data from the WISDOM project, a global initiative to use grid computing for drug discovery, and consisted of 750 chemical dockings—potential candidates for diabetes drugs. In about 12 hours they ran calculations corresponding to 55 days on a single processing unit, Venton revealed.
“While this was small for us in comparison to previous sets of calculations,” Jean Salzemann, computing researcher with the WISDOM collaboration, said in the article. “It tested new ways of sending jobs. When we are preparing future data challenges, we could consider splitting the data load between EGEE and Digital Ribbon resources.”
In its latest calculation, WISDOM tested 4 million potential drug candidates, a task that would have taken a single computing core four centuries to complete. WISDOM achieved the results in two months using 5,000 processing cores.
Related posts:
Category: Applications, Networks
Follow us...
If you enjoyed this article, connect with us to receive more just like it.
