For the past decade as a software researcher, technologist, and data scientist, I have watched the movement toward the next “big thing” in computing from the days of “grid computing,” which promised to create virtual organizations of computing, resources, and people, to today’s “cloud computing,” which is admittedly quite different than where I believe technologists decades ago imagined we’d be in terms of computation, scalability, privacy, and interconnection.
Cloud computing was widely popularized during the mid 2000s, as internet properties including Google, Amazon, and others found that the clusters of commodity computers they were running their highly scalable businesses from could be further virtualized and made available internally (and eventually externally).
These companies moved away from the traditional model of buying exotic “big iron” hardware with typically large support contracts from vendors, and moved into the realm of buying cheap, easily replaceable “commodity” parts, reducing the specialization and skill required to maintain and evolve the hardware and capability.
What is Cloud Computing?
What exactly is cloud computing? According to the National Institutes of Standards and Technology (NIST), cloud computing must satisfy five core principles, including providing on-demand service, broad network access, resource pooling, rapid elasticity, and measured service.
In short, it must provide computing that allows businesses to scale dynamically, at measured cost and value, and provide broad access to a greater pool of resources than would be cost effective to create yourself.
In the early days of cloud computing, there were public, private, community, and hybrid clouds. Public clouds include computing resources available for purchase from companies such as Amazon or Google. Private clouds are created using internal company hardware and resources, and then offered using cloud application programming interfaces and tools. Less prevalent today are community clouds, where private clouds are shared by multiple corporations. Hybrid clouds use a mixture of private and public clouds to jointly deliver service.
On top of clouds, businesses deliver service in three main ways. The Infrastructure as a Service (IaaS) model provides low-level tools, such as servers, networking hardware, and storage, for software developers to use as building blocks for scalable and elastic cloud resources. Examples of IaaS include Amazon’s S3 storage service and its EC2 service for elastic computing.
The Platform as a Service (PaaS) model provides more structured services and higher-level tools to build apps, an example being Google’s App Engine.
Finally, the Software as a Service (SaaS) model provides more traditional user tools delivered in a cloud environment, for example Google’s Gmail for electronic mail and Docs for authoring documents. Today, businesses invest 48 percent of their cloud computing IT budgets in SaaS, 30 percent in IaaS, and 21 percent in PaaS.
Where is it Being Used?
Since many businesses only need increased computational and storage resources seasonally—think banks and government during tax time, or retail businesses during Christmas or other holidays—small companies can sometimes get by with private clouds. But for medium to larger companies who do greater business, there is increased reliance on big cloud providers. Because of this, 77 percent of businesses today use cloud for enterprise services.
According to a study by IDG publications, SaaS is the most widely used cloud service delivery method. This is likely due to the pervasiveness and ease of delivery of simple applications related to productivity, such as email, to a big user base.
According to the IDG study, telecoms companies are set to be large users of cloud computing services, which makes sense due to the growth of today’s Internet of Things (IoT) small devices and connected platforms. Projections indicate there will be 20.8 billion IoT devices by 2020.
IoT devices include popular voice-command tools, such as Amazon’s Echo and Alexa, and Google’s Google Home appliance. These voice-controlled devices listen to your wants and needs and then issue commands that kick off cloud processing. The results might be sent to networked devices in your home, including light switches, garage doors, power plugs, and smart doorbells, such as the Ring product.
These IoT devices typically ship with little to no computational power (though that is changing), and instead rely on the cloud for beefy processing, e.g., finding an intruder in a doorbell video; looking up what the weather will be, or what your energy needs are and dimming the lights, etc.
Promising networking and infrastructure improvements, including the movement towards the 5G phone network and Bluetooth 5.0, will allow for millions of connected IoT devices with a bandwidth resembling the powerful and fast Wi-Fi we’ve come to expect from desktop computing.
Many industries rely on cloud computing for their business needs today—banking, health care, insurance, energy, and utilities to name a few. In the realm of banking, Capital One’s mobile banking now relies primarily on Amazon’s Amazon Web Services (AWS) IaaS offering.
In the domain of health care, there is increased use of cloud and SaaS for electronic medical records and electronic health care record systems, and also for productivity tools such as Office 365 for email, document authoring, and so on. In the insurance industry, cloud computing is clearly a big business driver, but less clear are specific examples of its use thus far. Finally, in the energy sector, with the emergence of “smart grids” the global Smart Grid as a Service market is projected to be worth as much as $6 billion by 2025.
Within the U.S. government, spending on private cloud computing was $1.7 billion in the 2014 financial year, and its use has grown significantly since then. Government has driven specific cloud offerings and usages, including the Amazon “Gov Cloud,” which ensures that cloud resources for computing, storage, etc., are provided on U.S. computers and compatible with U.S. International Traffic and Arms Regulations.
Government use of clouds is tempered by the need for applications to be certified as compliant and have appropriate security under the Federal Information Security Modernization Act, Federal Information Processing Standards, and the Federal Risk and Authorization Management Program. This set of laws and regulations helps to ensure that applications built in government-used clouds do not compromise national security, or allow threats from both inside and outside the government.
Largely, cloud computing has been a resounding success. Considering that one of the major cloud providers, Amazon’s AWS, had an unimaginably low downtime of just 2 hours, 30 minutes for the entire year in 2015, service offerings are responding to the reliability needs and demands of businesses across a variety of industries.
However, there has been nothing to change that, in reality, there are four major providers—Amazon, Microsoft, Google, and IBM—that combined account for about two-thirds of all cloud-based storage, and so finding other enterprise options besides those four is not currently possible.
Where is the Cloud Headed?
With the increasing sales and deployment of IoT devices, and the increase in network capacity, including Bluetooth 5.0 and 5G cellular networks, there is a big movement towards “edge computing” or computing on the previously non-compute-heavy edge nodes in the IoT network. The idea behind this is that powerful artificial intelligence and machine learning needs to run on these devices to do things like detect an intruder, or to contextualize your preferences and make recommendations for you.
This is also called “fog computing,” coined by Cisco in 2014. Moving computing to these IoT devices is also something widely considered to support the emerging trend in self-driving cars and autonomous driving and vehicles.
It’s an exciting time for deployment and utilization of cloud computing across the world, especially considering the emerging networking upgrades and wide ability for business growth and use.
Chris Mattmann is a principal data scientist and associate chief technology and innovation officer in the Office of the Chief Information Officer at the Jet Propulsion Laboratory in Pasadena, California.
Views expressed in this article are the opinions of the author and do not necessarily reflect the views of The Epoch Times.