Up With Virtual Grid Power

I recently wrote a column for InformationWeek Analytics that got some e-mail responses, and I thought the discussion was interesting enough to post the column and some of the comments that sparked the discussion. So here goes.The Newsletter Column

In this space last week we discussed the problems industry leaders have with the term "cloud computing." But terminology is the least of this concept's problems. The big issue, as I see it, is that this buzzword came out of nowhere and introduces nothing new whatsoever. It still requires the same type of hardware and software, does not solve any new challenges, and does not introduce a new way of doing things.

Think Salesforce.com is cloud computing? Nope. The idea of a hosted application that works over the Web instead of having a client/server model has a name-or a few, actually. Remember ASPs? How about SaaS? OK, so you say hosted processing power is the cloud. We can now pay for what we need as far as storage and virtual machines. Been there, done that, got the tee shirt.

Here's what WOULD be innovative: Using new virtualization software in conjunction with grid computing. That's a concept that should get more attention. We could finally build applications in the "cloud" that can draw their computing power from several different computers on the network or over the Internet. Take all the desktop machines that you likely have in your company. After 5:00 p.m., most employees go home, and you have all this processor power that is available, sitting there collecting dust. If we could somehow harness these computers to fuel faster processing of applications that could lead to faster task completions, that would be cool.

What if a virtual machine, for example, lived in the "real cloud" that I'm describing. It would draw its computing power from several hundred machines, and it would not be bound to any host, so the potential for this VM to stop working due to a host failure is "virtually" impossible. Some of you may know seti@home, an application that uses the computing power of several million PCs around the world to analyze radio signals, searching for signs of extraterrestrial intelligence. Something like that on a smaller scale would spark my interest in this buzzword.

The real harm here is that the noise around tacking a new term onto old ideas is a distraction from true innovation. If you disagree, I would love to hear from you.

Reader comments Daniel Longo wrote to me: Elias:

Enjoyed your post on virtual grid power. You only used acronyms when appropriate, and you didn't use the word 'product' once. It seems 'cloud' has become just a new word for salespeople to use (my background). We couldn't sell you ASP (or is it the ASP product?), or SaaS, so maybe you'll buy some cloud. At least it's not another 3 letter acronym. Good explanation for your interesting idea for unused processing power. I'm sure there are some issues with licensing and security, but there should be a way to make it work. The idea seems well suited to certain tasks like global updates, end of month accounting or maybe e-discovery.