Cloud Computing & Rules of Machines in the 21st Century

Dr. Hossein Eslambolchi
March 2012

Cloud computing is a model that enables ubiquitous, convenient, and on-demand network access to a shared pool of configurable computing resources. These resources can include networks, servers, storage, applications and other services. Cloud computing allows these resources to be rapidly provisioned and released with minimal management or interaction with the service provider.

Cloud computing, combined with the processing power of quantum computing, will create a future familiar to science fiction fans: Ubiquitous, powerful, and intelligent robotics will appear by the end of this decade, and will go on to dominate our world by 2025.

In this decade, however, cloud computing is becoming inextricably linked with our social, political and financial lives. We need to understand the profound effect cloud computing is having on our lives in order to optimize our enterprises across the globe and within different vertical industries.

The growth of cloud computing is a force to be reckoned with. Because of its growth, I predict global traffic flow will increase from 130 Exabytes in 2011 to over 1.6 Zetabytes annually by 2015. Additionally, with the widespread use of video streaming, I would expect to have over 22 trillion hours of streaming video available for viewing by 2014. 5 trillion of those hours will be dedicated to business webcam conferencing alone. It is also clear that with the advent of 3D TVs and bandwidth-hogging 3D content, more than 1.6 trillion hours of on-line HD/3D video streaming will be available and delivered over the Internet by 2014.

Along with growth in IP traffic, we also need to consider the power of cloud computing in relation to data centers and their traffic-carrying load. Data center traffic is expected to grow at 33 percent annually, starting at 4.8 Zetabytes in 2011. Nothing can stop this growth; the number of applications and innovations which need server infrastructure and more bandwidth will drive it on. I also believe that 76 percent of IP data will stay within the data center as virtual machines migrate from one server to another server. The use of Cloud computing in this world becomes paramount in designing next generation data centers.

For example, today Netflix represents 33 percent of U.S. peak downstream traffic, and video makes up altogether 32.6 percent of peak downstream mobile traffic. YouTube is currently the largest contributor, but other applications like live sports will also add to the overall flow.

Consider the massive storage capabilities of the future and advanced analytics like linear programming for networks, and it’s easy to see that we will be dealing with massive amounts of data. Therefore, knowledge mining will drive new businesses and new solutions through the use of analytics and deep packet content information.

Finally, the total economic value of cloud computing will reach more than $40 billion in 2012 and rise to more than $100 billion in 2015. Cloud computing unleashes the a level of utility computing that has never been seen before. This is a game changer, especially when matched with advanced communications and applications.

 

RULES OF THE MACHINES & CLOUD

Back in the 1940s, we used to perform computations at a rate of 1 computation per 150 seconds. IBM SSEC produced around 500 cycles per second (CPS). Forty years later, in 1980, we enjoyed improved computational power, with IBM’s PC reaching more than 250,000 CPS. The Cray Computer ran at 86 million CPS for scientific and heavy analytics applications. Thirty years later, the Earth Simulator hit 38 trillion CPS, the ubiquitous Sony PlayStation placed 2.1 trillion CPS in our homes, and Apple’s iPad allowed us to hold a 1.7 billion CPS computer in our hands.

This progress is amazing, but I predict that nanotechnology could create a computer that performs 8.6 quadrillion CPS. When quantum computing finally arrives sometime around 2016-2020, the amount of computation will be so massive that technologies that once seemed like science-fiction will become easy to implement in the real world.

The life of everyone on this planet will change over the next decade. Along with the triumph of IP worldwide, we may also see similar irreversible and global revolutions in virtual reality and robotics. This prediction may seem like a stretch, but I can envision it quite clearly when I consider the massive amount of computing power nanotechnologies that will become available in the next few years.

In short, computing will surpass Moore’s law by a wide margin, and Moore’s law will become history, just like all other laws of physics have been surpassed by new insights.

The uses of this massive computing power are, frankly, mind-boggling.

Think about the present. We make massive investments in telescopes that allow us to see 10 billion light years away, the greatest distance in the Hubble Deep field. (In 1950 we could only see 10 million light years from the Earth.)

But the new technologies I’ve just discussed will allow us to do so much more. Gamma Ray technology and quantum computing may soon allow us to see more than 12 billion light years away. A revolution in astrophysics will follow; we may finally be able to “see” the earliest traces of the big bang, or definitive proof of life outside our solar system.

The point behind all this speculation is that computation will soon become so fast, and the data at our disposal so massive, that a revolution will bloom in every aspect of our lives iIncluding business.

This is the reason why an early embrace of cloud computing is critical for the survival and success of companies in the 21st century.

There are some key issues that service providers must confront in order to ride this massive transformational wave. Cloud computing will become a $100 billion market by 2015. Let me share with you some of the important issues, in priority order, that we all will confront in order to harness this amazing transformation.

  1. Security and Availability
  2. Unified Communications
  3. Enterprise Applications
  4. Cloud Orchestration
  5. Cloud-Based Services
  6. Data Analytics
  7. Converged Services
  8. Infrastructure Value
  9. Platform Virtualization
  10. Network Virtualization

In my upcoming blog, I will explain each of these issues in more detail for the benefit of network engineering and design.