While at OSCON in Portland recently, I took part in a panel alongside Rishidot Research founder and principal analyst Krishnan Subramanian and TechCrunch writer Alex Williams – the panel was an attempt to get some industry observers together to discuss the future of the cloud. Krish has written about the session here at Cloudave. In his post he discusses his criticism for solutions that require specific combinations of hardware and software. In his post, Krish says that:
I am against bundling hardware with software layer and call it a cloud… If cloud is about abstracting away all the hardware complexities underneath, I shouldn’t be forced to buy a specific hardware to run the cloud software. Well, you can always make a point that these hardware+software solution will help organizations build clouds. Yes, they do but their offerings by itself is not cloud. Period.
Now I don’t disagree with Krish’s contention that tightly specifying hardware and software combinations and packaging them up as a “cloud” is more marketing than substance, but I suspect my reasons for thinking this are somewhat different than his. When answering the same question from Alex, I agreed with Krish’s perspective, to a point, but added that talking about cloud as merely a technology kind of misses the point.
My belief, and it’s one I’ve espoused before, is that cloud computing is all about taking technology solutions and delivering them in a way that allows organizations to consume them “as a service.” I’m happy with the concept of private cloud, with cloud in a box or any other permutations of “cloudiness” so long as it means that organizations can give their users the ability to enjoy the sort of flexibility that we, as consumers, enjoy with services like Gmail.
What does this mean in the real world? Well if I’m working in the marketing department of a large enterprise, I should be able to spin up some servers for a short term campaign. I should be able to do so without IT intervention. I should be able to spin them down at the end of the campaign. I should receive a bill (even if it’s merely internal cost allocation) for the services I’ve used. ideally, I should have visibility over both the expense of that service, and the financial outcomes for the project.
Quite simply, cloud is about delivering business agility, flexibility and transparency. If that can be delivered from a tightly packaged hardware and software solution, then so be it. Of course, I have my own views on long term trends and, as far as I’m concerned, the future will see a gradual shift from on-premise, to private cloud, to public cloud – that’s a factor of economics and scale. But the bottom line is that cloud isn’t about meeting a series of technical check boxes; it’s about meeting the needs of the business.
Related articles
- Hardware, Cloud And Cloud Washing (cloudave.com)

(Cross-posted @ The Diversity Blog – SaaS, Cloud & Business Strategy)
you seem to be contradicting your own point. you are saying that you are ok with software coupled with hardware as long as it offers flexibility. then you are arguing that marketing department should be able to get the services without IT. tell me how marketing department can get services without IT, if it is buying software coupled with hardware. are you sure you are clear on what you are trying to articulate here? do you want to talk about flexibility of business user or IT? i get a feeling that you have no enterprise experience.
John – the whole point is that with good systems in place (management, visibility, chargeback) corporate IT can put in place some internal cloud-like infrastructure that business units can then utilize as they require. It allows them to spin up servers essentially at will, albeit that those virtual servers sit on physical on-prem hardware
Flexibility and scalability. The best, strategic cloud deployments should empower an organization to seamlessly rev up or down depending upon time sensitive needs. This is true whether the organization needs cloud processing for R&D or depends upon cloud-based collaboration solutions (i.e. Avaya’s Live Connect) to power communications.