As I promised a few weeks back, GRIDtoday will be providing some more in-depth coverage of the Media Grid project. And, this week, it’s here.
I know a lot of different people have a lot of different thoughts about utility computing, and there are valid arguments both for and against it, but if the minds behind the Media Grid succeed in their ambitions, we could see some of those perceptions start to change. The project is unique for many reasons, not the least of which are its dedication to open standards and an economic model that rewards resource donators, but the most unique aspect might be that it already has at least one user committed to signing on as soon as the grid is ready for public consumption. And Vertex Pharmaceuticals seems to me like a company with too much to risk to jump into a new way of doing business without believing wholeheartedly that it will be a positive experience.
After all, as Vertex’s Paul Dupuis told me, drug development is an “extremely long, involved and expensive process,” and moving from an in-house cluster to an open, transparent utility grid is quite a leap in a business where time not only means money, but lives. Vertex already has developed treatments for HIV/AIDS, and currently is working on treatments for Hepatitis C and cystic fibrosis. But, because he believes strongly the Media Grid’s economic model, and because he is tired of upgrading and being tied to one vendor, Dupuis is ready to go once the Media Grid gets its standards adopted.
It’s one thing to hear about some boutique graphical design firm using a utility computing model to handle it’s occasional need for HPC, but to hear of a pharmaceutical company prepared to all but migrate its entire HPC infrastructure to a utility model is another story. To the doubters, it has to inspire a little confidence …
Another must-read article this week comes from Ian Foster, who originally wrote “IT as Fuel for the Innovation Engine” as part of GRIDtoday and HPCwire’s special coverage of last week’s LinuxWorld event in San Francisco. As usual, Foster does a great job of laying out the case for Grid computing as a primary tool for innovation, utilizing a hypothetical situation based on his work with an actual company. As for the LinuxWorld coverage, you can get all three days worth of breaking news and special features by visiting www.taborcommunications.com/hpcwire/linuxworld/index.html. Aside from Foster, the coverage includes articles by and Q&As with leading folks in the fields of HPC and virtualization.