What you might think the cloud is actually isn’t what the cloud is. Worse, a new survey commissioned by Citrix finds that many people don’t know what the cloud is anyway, but what they think it might be isn’t what it actually is. (Yes, this is meant to be confusing).
Let’s break it down. Cloud-computing – the real cloud computing – was intended as a solution to a processing problem. “The promise of cloud computing was that you had vast resources and could get everything done off of your desktop – or smartphone or tablet,” explains Dan Coffey, an independent technology guru whose background includes stints at both Bell Labs and Kendall Square Research (he’s BillingViews’ go-to guy on the really complicated stuff…).
Turns out that when crunching massive processing loads, let’s say to simulate how atmospheric particles interact with a new orbital rocket design, it can be more efficient to distribute computation across a network of computers than to try to run the same computations on a single machine, even if that machine was loaded with processing power. The cloud aspect of it is that applications responsible for running those computations would leverage a network to identify and access available computing resources and would be responsible for managing the distributed processing and pulling the results back together in the end. The problem, says Coffey, is that “applications haven’t been re-written to take advantage of all of the processors out there.” Coffey adds that what large enterprises call the cloud “is really just linear code running on 1, 10, or 100 processors; you get volume, but not speed.” Further, Coffey says, “we’re kind of maxed on processors and have been for two or three years. So, you go to two, four, or six cores, but the applications can’t take advantage of them.”
So, since what was the cloud hasn’t really come to fruition, what we we now call the cloud is essentially hosted data centers. We’ve just had an evolution of terms from service bureau, to hosted service, to ASP, to SaaS, and now generically “cloud.” So what’s the difference?
Perhaps the most important change from the days of the service bureau to today’s potentially misnamed “cloud” applications has been the rise of the web services-oriented API. Because of today’s APIs, users get the benefit of flexibility and integration at a reasonable cost and shed the overhead of hosting, licensing, and maintaining their own applications. It’s not as rosy as all that, but arguably today’s APIs are what enable the concept of reaching into “the cloud” – one of many data centers that host applications you want to tap into – for just the data or functionality you need without the overhead of integrating yesterday’s heavy-duty APIs or managing major on-site implementations.
So, that has almost nothing to do with the original idea of “cloud computing.” Hence, what cloud is, what we now refer to as the cloud, and what people believe the cloud to be are all different things. The only thing we know for sure is that a cloud absolutely is one of those white puffy things in the sky; what you believe its shape resembles is also open to subjective interpretation.