Grid Computing – Distributed processing & storage for TV

Faster, Bigger, More Reliable, Cheaper. That's what grid computing promises large scale moving image users. So says an article in Broadcast Engineering. It is intriguing stuff and is clearly going to be a core methodology for designers of large digital media systems.

OK cheaper may not spring to mind readily to begin with, but there are huge down-the-line cost savings to be made and comparing a non-grid based infrastructure to a grid-based design there are some things that would be prohibitively expensive to achieve with the former, if they were possible at all in some cases.

So what is it? Not too hard to imagine really; a grid of resources that can be operated in parallel and whose outputs can be combined at the last moment.

Grid computing has been around for a long time as an idea. It has been used, too, for some time in power computing applications, but it is only now that the TV world has begun to have access to the technology.

At NAB this year, Omneon launched their new MediaGrid architechture bringing the speed and reliability increases to large scale users.

The article points to some useful reading material at  IBM's library of Redbooks. In particular there are a couple of pertinant documents worth taking a look at to begin with: “Fundamentals of Grid Computing” and “New to Grid computing”.