Since 1986 - Covering the Fastest Computers in the World and the People Who Run Them

Language Flags
August 22, 2008

Grid vs. Cloud vs. What Really Matters

Derrick Harris

Is anyone else just plain sick and tired of bloggers trying to decipher the difference between cloud computing and grid computing? I think every other one of my news alerts is on this subject. Don’t get me wrong, I’m as guilty as the next guy when it comes to arguing over semantics, but, at this point, there doesn’t even seem to be room for confusion when it comes to these two computing paradigms — at least when we’re talking connotation.

Yes, there are some similarities (indeed, fair degree of derivation grid –> cloud) underlying technology-wise — load-balancing, scheduling, parallelism, prioritization, etc. — but these are all but mooted once the discussion moves to use cases and overall perception. Just ask a CIO. Grid computing = high performance, one application, possibly batch-scheduled. Cloud computing = often transactional, a variety of services, on-demand — 24×7. Are there service-oriented and transactional grids that blur the line? Yes. Are there clouds (e.g., the current carnation of Network.com) that focus on high-performance batch jobs? Of course. But at the end of the day, the discussion most often is framed in this context. With the exceptions of EC2 and Network.com, there are very few (dare I say no) clouds on which you would build, test or run compute-intensive applications. And it would seem equally ridiculous to host a Web application on a traditional grid infrastructure.

I realize this an oversimplification (and doesn’t even address SaaS-style cloud computing), but to the extent we must distinguish between the terms, I think most folks would agree that the connotations above mirror the perception of most IT users and analysts.

Of course, we’re supposed to ignore buzzwords and just focus on the business benefits anyway, right? We’ll be publishing a piece shortly wherein old-guard grid vendors address just this issue. The reality is that most seem to be heeding analyst advice by broadening their product portfolios, lessening up on the grid talk, and reserving the cloud discussion until customers bring it up. Believe it or not, they’re simply talking about what the technology means for the company.

But grid versus cloud is hardly the only terminology confusion spreading across the IT landscape. Just this week, we ran two articles talking about database solutions — one calling itself data caching and the other calling itself database virtualization, and both using, to some degree, the term data grid — and there are more to come. Are they data grids, are they virtualization, are they caching? They’re probably all of the above, but the proprietors aren’t too concerned with what you call them. If you need continuous availability, Gridscale might be the right choice. If you need real-time data access, Coherence might be more up your alley. Like with grid or cloud, customers need to start with figuring out problems they want solved, not what technology they want to bring in, and the right solution should surface.

And have you seen James Urquhart’s “Cloud Computing Bill of Rights?” I can’t really argue with what he has come up with thus far — essentially, openness and honesty about what they’re offering on the vendor side, and due diligence and best practices on the user side. Like our country’s Bill of Rights, it’s the relationship between the two sides that makes everything work.