Opinion: By using its expertise in backing the World Community Grid project, IBM also gets the chance to demonstrate the benefits of grid computing.
Making use of unused CPU cycles on your client computers isnt a new idea. Going back into the 1980s, there was network database software that let you install a small piece of agent software on your client computers that, after normal business hours, would allow the database server to distribute its indexing load to any computer that was running its agent. In the early 1990s, graphics software was developed that, using the same agent model, was able to distribute the image rendering process to many different types of client operating systems, speeding up what is still a very CPU-intensive process, rendering graphic images.
But the end of the 20th century saw not only a massive increase in the number of network computers, but also freely distributable client software that worked together with a centralized server to complete a specific task.
These clients, such as the RSA encryption cracking contest tool from Distributed.net
and the search for extraterrestrial intelligence from SETI@Home
, provided clients for just about every common client operating system, let the user determine how much CPU resource they would use and when the software would run, and gave users a sense of camaraderie in creating teams that competed to devote the greatest number of excess computing cycles to the selected project.
Now IBM has taken this concept a step further by stepping up as the technical muscle behind the World Community Grid project
, joining United Devices (the folks behind SETI@Home) and a host of academic and scientific organizations to create an organization that uses these spare CPU cycles to work on projects designed to benefit humanity.
Click here to read why IBMs database VP says IBMs grid scales better than rival Oracle.
While these distributed computing projects are clearly not new, this is the first time a corporation of IBMs scope has taken an active role in such a broadly targeted project.
The WCGs first project is the Human Proteome Folding Project, which is designed to determine how the proteins coded by the human gene sequence are likely to fold. This information can help researchers develop new disease treatments. Expecting tens of thousands of end users to sign up, IBM is putting some serious server hardware behind the effort as well as its Shark Enterprise Storage Server and a DB2 database. IBM is also suggesting that its employees sign on their computers as clients for the WCG projects.
Future projects will address what the advisory board of the WCG calls "important global issues" at the rate of five to six projects per year. The organization is actively looking for ideas, suggestions and proposals for future projects, and the WCG Web site has a downloadable RFP form in case you have a project that you think will catch its interest.
As altruistic as this all sounds, there are sound business reasons for IBMs support of the WCG. From its current leadership position in the grid computing market, its perfectly clear that the company needs to keep evangelizing all of the concepts that make up practical grid computing applications, and the distributed client model is certainly a valid one. For this reason, the WCG should be considered a technology demonstration for IBMan ongoing work that continually demonstrates the practicality, reliability and security of a large-scale distributed grid implementation.
Check out eWEEK.coms for the latest utility computing news, reviews and analysis.