IBM 'Clouds' Look Like Conventional IT

According to this e-Week report, and this report in the New York Times, IBM continues to form its cloud computing strategy, including the definition of some key products.

IBM will announce a number of new cloud computing products and services... In addition, IBM will unveil plans for a new research lab focused on cloud computing. The initial plans will be for a bundle of hardware, software and services aimed at software developers and testers, and another bundle targeting virtual desktop environments. The moves come at a time when every major IT player, including HP, Cisco, VMware, Microsoft, Google and Amazon, are making deeper in-roads into the cloud.

The issue here is that cloud computing is really about, well, cloud computing. Existing hardware and software vendors, including Microsoft, Cisco, HP, etc., and of course IBM, seem to find that thought a bit scary and continue to toss traditional hardware and software at the problem.Want to move to cloud computing? Do so as a private cloud first and see how it goes, according to IBM. By the way, you're going to need a few million dollars in new hardware and software to get you down the road to the cloud. That seems a bit counterproductive, if you ask me.

The trouble with cloud computing and other paradigm shifts, including SOA, EAI, component-based development, etc., is that those with the marketing dollars typically need to push hardware and software on the masses and use these shifts as mechanisms to push product. However, cloud computing is a bit different. It's about using less hardware and software, not more, and that means the hardware and software vendors need to restructure their marketing efforts to redefine just what cloud computing is, such as "private clouds" and "hybrid clouds," rather than simply drive everyone to Upgrade and Replacement City. Don't get me wrong, you need private and hybrid clouds within some architectures, but their use should be rare in the context of the larger cloud computing movement.

Cloud computing is not about hardware and software, it's about leveraging computing resources that you don't own that may, and I repeat 'may,' be able to provide you with a much more economic, elastic, and scalable way to provide enterprise computing. If you begin to shop for new hardware and software, you've already lost the war. Chances are you're just adding risk and cost, and not gaining any efficiency.

The larger danger here is what I call the hype hangover. As we continue to move toward cloud computing, the expectations being set by the vendors also setup cloud computing for failure. They are selling a commodity and therefore tend to define the commodity in ways that will best meet their sales and marketing goals. We can't get there by expanding our data centers. Cloud computing is about doing more with less, and not about another round of software and servers. We've been there before, and it is not the road to the key goals of cloud computing: Economic and scalable efficiency.The issue here is that cloud computing is really about, well, cloud computing. Existing hardware and software vendors, including Microsoft, Cisco, HP, and, of course, IBM seem to find that thought a bit scary and continue to toss traditional hardware and software at the problem...