Better, Faster Infrastructure: The Roots and Effects of the Cloud Revolution
Wednesday, November 20, 2013
Posted by: John McDonald, CEO CloudOne
Editor's note: This article was contributed by John McDonald, Chairman, President and Chief Executive Officer of CloudOne, who is a member of the East Central Indiana Chapter. It was a paper presented at a recent IBM conference.
For a PDF copy of this paper, see: http://www.aitp.org/resource/resmgr/2013-ie-files/cloud-revolution.pdf.
AITP encourages other members to submit articles for consideration to IEeditor@aitp.org. This helps us fulfill our collective mission to share IT knowledge among our members.
Abstract - The adoption of cloud computing is not a
technical revolution, but instead is effectively a procurement revolution, led
by the desire for economic efficiencies in the deployment of infrastructure and
collaboration of skilled workers.
NOT A TECHNICAL
Speaking at the MIT Centennial event in 1961, computer
scientist John McCarthy, creator of the term "artificial intelligence” and winner
of the Turing Award, Kyoto Prize and National Medal of Science, said:
may some day be organized as a public utility just as the telephone system is a
public utility... The computer utility could become the basis of a new and
Indeed, as far back as the 1950’s, organizations practiced
"time sharing”, a popular method of spreading the cost of expensive computing
power across multiple teams. However, limitations on data transfer speeds
through public telephone networks limited the practice to highly localized
pockets of computing power, such as at universities or corporate campuses.
No single technology nor any single moment removed these
limitations. Instead, in the past decade a group of technologies have converged
to create the conditions that allow for computing resources to be shared
broadly throughout the world. In April of 2008, Galen Gruman and Eric Knorr,
writing for InfoWorld 2,
outlined these components as:
Networks. These tie together computing resources throughout the world
through the Internet.
Computers and Storage. These offer
the ability to provision large data centers at historically low capital costs.
Virtualization. Allows for the time-slicing and sharing of processing power
into sub-machine virtual machine images.
Architecture. Breaks up large computing tasks into loosely-coupled
services, connected by open standard communication protocols.
Computing. The ability of computers to self-manage and "selfheal” from
common problems, making them more reliable in clustered configurations, as no
one device provides a single point of failure.
The combined effect of these technologies is to enable the
creation of centralized "super centers” of computing, accessible through
high-speed Internet connections, stocked with dense clusters of self-healing
computing processing and storage equipment, running autonomous "virtual
machine” computers. On those computers are portions of software "stacks”, such
as operating systems, databases, web application servers and application
software, linked together by open-standards communication protocols in a
service-oriented architecture. The virtual images are transportable from one
storage and/or processor to another, within the same center or to a
geographically distributed center for backup or disaster resiliency. We call
this combined architecture cloud computing.
These cloud computing centers are effectively the "power
plants” of a utility computing model, just as Dr. McCarthy envisioned in his
speech at MIT. The Internet serves as the high-voltage, long-distance power
transmission lines, while local Internet service providers represent local
power utilities in the metaphor, connecting individual homes and businesses to
the long-distance transmission network, and billing and measuring for service
Therefore it is not one inventor or invention that created
the cloud revolution. It is the convergence of the combined evolution of a list
of computing-related technologies that is responsible for the emergence of the
IT IS A
In the past, Information Technology leadership spread
computing resources closer to users, as linking geographical-distributed users
to centralized computing resources was expensive and frustrating. Unfortunately, this simply shifted the
expense and frustration to the teams of people responsible for maintaining,
repairing and supporting those resources, forcing them to adopt their own
complex and expensive methods of performing their tasks.
This created a bit of a "serfdom” of users beholden to those
who held the resources and skills to maintain the computing resources required
for business functions. For example, complicated methods of prioritizing the
provisioning of new servers developed, meaning that no matter how urgent your
need for capacity or assistance in your business unit, your requests were
prioritized and weighted against all other requests through a central "command
economy.” This potentially resulted in delays, extra costs, and frustration if
your needs were deemed to be less important than others.
As noted by Teresa Takai, Chief Information Officer of the
United States Department of Defense in her July 2012 Cloud Computing Strategy
document, the emergence of "power plant” cloud computing centers is enabling a rapid
consolidation of computing resources, and with them the skills to maintain
This enables lines of business to more freely choose
alternative providers of computing power and skills, instead of being beholden
to a single internal source for all. This is accelerated by the recent trend to
push budgetary authority and profit/loss responsibility to leaders in of those lines
The result is a tremendous pressure on Information
Technology leadership, forcing them to evaluate a shift from being the source
of all computing power to being the partner in procuring the best of breed
technology for business function. In their report, On Demand: From Capacity to Capability, Gartner researchers wrote:
are experimenting successfully with internally and externally sourced capacity
on demand. Within 10 years, they and their enterprises will be able to acquire
complex outcomes— capabilities—on demand. This will dramatically change
enterprise IT organizations and the software industry that serves them.4
This "democratization” of computing power, placing decisions
in the hands of lines-of-business, is the true, core effect of cloud computing.
It is not a technological revolution, but instead is a procurement revolution.
EFFICIENCY AND GLOBAL COLLABORATION
Two strong market forces motivate this democratization of
computing power: efficiency and the need for global collaboration.
Efficiency is the pursuit of maximum benefits from minimum
expenditure of resources. In the cloud computing space this takes two forms: specialization
of skills and elastic adaptation to changing business needs.
In regard to skill specialization, organizations are realizing
that in many cases the construction and maintenance of computing resources is
not a core competency. They are instead seeking vendors and supplier
relationships to offer computing resources on demand, tailored specifically to their
This concept is not new: organizations have effectively
"outsourced” many non-core business functions, such as real estate, payroll, administrative
staffing, and more recently office supplies and other business equipment such
as copiers and printers. Now the ability to get computing power on demand, in
nearly any location, specifically designed to individual needs, is enabling the
same "micro outsourcing” of processing, storage and software.
This leads to the next point of efficiency: elastic adjustment
of computing power as a direct reflection of usage. Historically, computing infrastructure
was sized based on the peak-anticipated load: an attempt to foresee the maximum
level of future need and "stockpile” enough capacity to meet that need should
it arise. Unfortunately this was essentially an educated guess, and was almost
never correct, resulting in shortages of capacity or extreme levels of waste, especially
for tasks with cyclical usage patterns, such as software and systems
The atomic nature of cloud computing resources allow for
micro-adjustments in capacity – storage, bandwidth, software images – so that alongside
automated provisioning and autonomic responsiveness, computing power can be provided
in lock-step with increased demands, and scaled down just as quickly as needs diminish.
This squeezes out inefficiencies in the old "stockpile” method, allowing
capacity for diminishing workloads to be automatically redeployed against
workloads that are increasing, effectively creating a super-efficient
marketplace of computing power in a cyber version of Adam Smith’s "invisible
In short, the efficiency driver of cloud computing comes
simply from the application of free market economics to computing capacity.
The second driver, global collaboration, also has its roots
in market economics. The lack of ubiquitous, low-cost communication technology frequently
necessitated the creation of pockets of skills, positioned close to their other
collaborators. For example, companies supplying components to a factory often
needed to create a nearby local production facility of their own in order to
ensure the free-flow of information and quick adjustments to changing workloads.
This in turn drove a "balkanization” of skilled resources into small widely
The same high-capacity networks have enabled these
geographically distributed teams to begin collaborating over long distances in
ways never before experienced in global history, as noted by Thomas Friedman:
before in the history of the planet have so many people – on their own – had
the ability to find so much information about so many things.7
Organizations are now creating centers of competency that
service everyone from a smaller number of organizational units, while at the
same time alternative work arrangements such as telecommuting and home office
workers are increasing the number of physical locations that need to connect to
the organization’s network.
This, in turn, is driving the need to centralize tools, data
and information into mutually sharable, common collaboration centers where everyone
can access the resources equally. Cloud computing fits this bill perfectly,
allowing for cloud-based resources that are shared with all, both inside or
outside an organization.
This external sharing enables a deeper relationship between
vendor and supplier, inviting point-to-point collaborations to speed products
to market, reducing cost, and allowing for even more specialized skills
procurement on the open market.
Therefore, the cloud computing procurement revolution is all
about economics: efficiency in resources and skills, and collaboration on a global
scale. These forces transcend technology, and are at the heart of human
Ever since early computer scientists such as John McCarthy
envisioned utility computing, the global technological infrastructure has
struggled to live up to the goal. The convergence of the Internet, low-cost
storage and processing, virtualization technology and software written to service-oriented
architectures has enabled the original vision on a grand scale. Organizations of
all sizes are adopting cloud computing not because of these technologies, but
because of the powerful economic forces of efficiency and skill collaboration.
This is forcing a procurement revolution inside of these organizations, which are
turning away from the old methods of inflexible, single-source computing
resources towards on-demand resources tailored specifically to their needs.
Ultimately cloud computing rests on the desire for better, faster infrastructure.
 Garfinkel, Simson (1999). In Abelson, Hal. Architects of
the Information Society, Thirty-Five
Years of the Laboratory for Computer Science at MIT. MIT Press. ISBN
 Gruman, Galen (2008-04-07). "What cloud computing
really means". InfoWorld.
 Takai, Teresa (2012-07). "Cloud Computing Strategy”. United States of America Department of
Defense. Public memorandum to secretaries of the military departments.
 Hunter, Richard. Iyengar, Partha. Rowsell-Jones, Andrew
(2010-12). "On Demand: From Capacity to Capability”. Gartner Executive Programs.
 Smith, Adam (1776). The
Wealth of Nations. Not in copyright.
 Van Alstyne, Marshall. Brynjolfsson, Erik (1997-03).
"Electronic Communities: Global Village or Cyberbalkans?”. MIT Sloan School, 1996.
 Friedman, Thomas (2005). The World Is Flat: A Brief History of the Twenty-First Century. Farrar,
Straus and Giroux. ISBN 978- 0374292881.