Grid computing on the wane or being incorporated

I've been writing articles for the IBM developerWorks Grid zone for many years, but IBM have decided to retire the Grid zone, citing the fact that so much software and many solutions now use many of the techniques originally pioneered in grid computing technology.

I don't agree with that - I think we still have a long way to go to get the best information and usefulness out of the technology. We are still yet to find a simple and effective method of managing and deploying an application over a standard grid solution (despite tools like Globus), and often those that do exist are overly complex for the type of wholesale deployment I think we need, or could take advantage of.

We still leave computers on overnight, and still prefer the model of a probably overpowered machine on the desktop and beefy servers in the background which could be altered and taken advantage of through grid technology. I've argued for that a couple of times before at least.

Interestingly, while looking around for more information, I found this piece at Computerworld: Grid computing: Term may fade, but features will live on.

Here they argue that grid technology is a solution for better accessing computing resources and that this is generally what we mean when talking about virtualization, software as a service and the service oriented architecture. While that's certainly a part of the process (one of my earliest articles was on using and exploiting grid technology with virtualization), but I think the fundamentals of making better use of our computing resources is still a noble goal, whatever you want to call it.

Is grid technology dead? I don't think so, and I don't agree that the term will fade yet either. Grid technology in it's original incarnation (massively parallel computing environments) will live on, whether you want to call that grid computing, distributed computing, massively parallel computing, or clustering.