24 Mar 2010 01:18 AM EDTposted by Michael Harries
Understand the underlying trends, but leverage the technologies
Cloud computing is about making enterprise quality technology available to every company. It's about reducing costs. It's about increasing agility. It's about consumer ease of use. It's about making computing into a utility. It's about outsourcing. It's about dynamic capacity.
Well yes, but ...
Cloud computing is best understood, not by talking about SaaS, PasS or IaaS or any other *aaS (footnote 1). Ultimately cloud computing is a transformation driven by several key technological shifts bringing about new ways to achieve old goals and enabling some new solution categories. It's a computing transformation that is only just beginning.
Cloud computing is best understood as a "phase shift" or industry-wide disruption, emerging from combinations of a small number of key technological factors. To explain, I'll use Professor W. Brian Arthur's framework from "The Nature of Technology". Arthur is a chaos scientist/economist noted for his seminal works "studying the impacts of positive feedback or increasing returns in economies, and how these increasing returns magnify small, random occurrences in the market place." (Wikipedia) These principles are especially significant in technology-specific industries. Arthur's most recent work "The Nature of Technology" looks at how technology changes over time; how it 'evolves'. Here's a small (paraphrased) portion of his framework.
- Technology: includes physical devices, processes, organizational structures. Technologies are overwhelmingly made up of other technologies as sub-components. The PC, for example, has various modular pieces of hardwire that are technologies in and of themselves, such as power supplies, hard drives, graphics cards, CPU, etc. Each of these again can be considered as having sub-components - less easily swapped, but sub-components (technologies) nonetheless that can be replaced at the design time (such as a new bus architecture for a CPU, etc).
- Domains: Technologies can be thought of as occurring in groups or domains - electronic, quantum, etc. This is related to the leveraging of common phenomena, but also to the nature of human specialization. Self similar types of components will be clustered. Some examples of this might be car mechanic, or computer engineer, or robotics engineer. Generally speaking technologies will be created from components that arise from skills/knowledge in a given domain.
- Technological change: does not occur like animal evolution - with random changes to the chromosome or by random combinations of pairs of technology characteristics. Rather, engineers or innovators will be aiming to achieve some goal and to improve the state of the art. Much of this change comes about by steady improvement in the components making up a given technology.
- Rebasing: Rapid change (Sometimes called disruptive change) occurs when solutions for a given problem are radically improved by replacing components of a technology with components from an entirely different domain. One example from computing is the way that the mainframe was replaced by the micro-computer. Relatively quickly the whole world of IT underwent a major change in cost, complexity and capability. Similarly, the airplane was transformed by the jet engine, the camera by digital recording.
Cloud Computing Drivers
Cloud computing fits the definition of rebasing described by Arthur. A number of major shifts (technology rebasing) are impacting the world of computing simultaneously. These include:
- The Internet
- Commoditization
- Virtualization (all flavors)
Let's look at these in turn.
- The Internet: we're still seeing the impact of the internet rippling through our computing landscape. Adapting to computing shifts can take time, and require levels of familiarity and trust to be built, as well as understanding the broader opportunities that arise. Internet cost models, architectures, and business models are becoming understood, but are still rapidly evolving. The ubiquity of the Internet is the major driver for cloud computing.
- Commoditization: is part of the natural process of the technology. In this case, two major commoditization factors affect cloud.
- IT as a business commodity: IT is no longer a business advantage, but a cost of entry. In other words, "IT doesn't matter" (Nicholas Carr (footnote 2)) Whether this is absolutely true is very arguable. However there is no argument that business IT is a maturing industry. In such a world, competition moves toward doing the same old things, but more efficiently. For example Software as a Service, by and large, is about addressing known jobs (such as sales force automation) with some changes in capability, flexibility and price. (Of course there are exceptions.)
- The x86/Windows monoculture: Intel and Microsoft have driven an extremely successful ecosystem of high volume, low cost chips and computers that have enabled an enormous part of our computing world today. "Monoculture produces great yields by utilizing plants' abilities to maximize growth under less pressure from other species and more uniform plant structure" Wikipedia. Monocultures are also susceptible to shocks. So the right (or wrong) push can have a huge impact across the whole monoculture. That is to say, the ubiquity of a given architecture raises the potential impact of certain technological innovations.
- Virtualization: including desktop virtualization, presentation virtualization, application virtualization, storage virtualization, server virtualization and many more. These technologies provide a way to separate one "level" of an environment from another; they provide an 'abstraction' layer. Abstraction and virtualization is a consistent theme through the history of computing - think about virtual memory, virtual users (in which each user appears to have a whole computer), virtual disks, and many, many more. While none of today's batch of virtualization is strictly new, the fact that they are being rapidly adopted by the (monoculture) market is very telling.
Together these factors lead to a dramatic re-basing of our computing environment, at an unprecedented scale. While cloud computing is the term du jour, it is representative of a shift in all of our computing toolkits.
Cloud is an ephemeral term, representative of a particular point in a larger scale rebasing of Information Technology to fully take advantage of the global Internet, a primed state of commoditization, and the broad adoption of a broad range of virtualization technologies.
Understand the underlying trends, but leverage the technologies
To anyone who makes it though this large blog post, I'd love to hear what you think,
Michael
__
Dr Michael Harries, Senior Director Strategy, Citrix Labs
@michaelharriesNOTES
1. This is seen in the overlapping and merging of these categories (SaaS, IaaS, PaaS), in the fact that much is missed (Desktop, Storage, Network, etc). Far more appropriate to move to terminology like 'software', 'platform' and 'infrastructure'.
2. It's not surprising that Nicholas Carr followed "Does IT Matter" with his cloud book "The Big Switch", given that the commoditization of IT is such a significant driver for "cloud".—
Finally, here's a great quote on Complexity Theory from W Brian Arthur:
"Complexity theory is really a movement of the sciences. Standard sciences tend to see the world as mechanistic. That sort of science puts things under a finer and finer microscope. In biology the investigations go from classifying organisms to functions of organisms, then organs themselves, then cells, and then organelles, right down to protein and enzymes, metabolic pathways, and DNA. This is finer and finer reductionist thinking. The movement that started complexity looks in the other direction. It's asking, how do things assemble themselves? How do patterns emerge from these interacting elements? Complexity is looking at interacting elements and asking how they form patterns and how the patterns unfold. It's important to point out that the patterns may never be finished. They're open-ended. In standard science this hit some things that most scientists have a negative reaction to. Science doesn't like perpetual novelty."
Extract from Wikipedia
This is relatively dense, but the message is simple. All that we are hearing with the word 'cloud' in it, is merely a shorthand for a particular point in the evolution of the Internet combined with a firestorm of virtualization technologies giving us a rapid shift in what's practical for mainstream adoption of IT.
