What cloud computing can and should be used for
[date-stamp]What can cloud computing do today, and what should it be used for? Since cloud computing describes a very generic resource, like a computer available anywhere, the versatility is its greatest advantage. But what can and should it actually be used for? File storage is a well-known cloud service and very basic, there must be something more it can be used for?
Flight Black Box
The idea of replacing the black boxes in planes with the cloud came from two former NTSB (US National Transportation Safety Board) employees in this article at gigaom.com. The black boxes, which are in all planes, contain a cockpit voice recorder and a flight data recorder. The idea and patent behind the black boxes used in planes today actually stems from the 1950s and 1960s. Professor James J. Ryan is credited with inventing it, and the US patent was granted in 1960. That's 54 years ago, and the world has developed a lot since then, so it may be time to replace this ancient technology with something more advanced. If flight data and voice recording can be transferred real-time to a cloud solution, this would eliminate the need to physically find the black boxes before their batteries run out (they send out a location signal for 30 days after a crash.) The investigation of a plane accident can also start right away when the data from the plane is available instantaneous. So no wonder the former NTSB employees floated this idea around-this would make the task of figuring out what went wrong in an accident much easier.
When a plane crashes, it is not only about finding the reason for the crash. Finding the plane and rescuing the people onboard the plane is the most important, and having flight data available right away when a plane goes missing is a tremendous advantage. Recently, Malaysia Airlines flight MH370 went missing, and days and weeks were spent trying to locate the site where the plane crashed. With flight data available from the cloud, finding the place where the plane disappeared would be the easy part.
The advantages to this approach are so obvious that the idea must be implemented, at least as an addendum to the existing black boxes.
The new game Titanfall for Xbox One (and other game consoles) is one of the first games to utilize cloud computing for some of the computation required for the game. Traditionally, graphics and advanced games have required a hefty game console (or PC), since all computation needed to be done locally on the game console. Not so with Titanfall-parts of the computation for the gameplay and graphics happen on Microsoft's Azure cloud computing platform. The results are transferred to the Xbox One console and shown to the user as a normal part of the game. Parts of the AI (artificial intelligence) in the game are computed this way.
In many ways, this removes some of the basis for requiring high-end hardware in game consoles. Granted, a decent graphics chip is still required, but a lot of the generic computation can be taken away from the game console and offloaded to the cloud. More games will adopt this clever technique. A reasonably fast internet connection is required though- South Africa was dropped from the initial list of countries where Titanfall would launch because of poor network performance.
Code tools for application developers
Visual Studio is a commonly used code editor, and this is now available as Visual Studio Online. Visual Studio Online runs on Microsoft's Azure cloud computing platform, and this rather demanding application is moved from developers' desktops and laptops to the cloud. Many spend hours every day in Visual Studio, so this is a big change for these people. There's no need anymore to have the most ultra-spec'ed laptop in order to run Visual Studio; the need for the computing power is moved to the cloud. Many developers will get high-end hardware anyway though, but at least this will now be for other reasons.
Applications are in general becoming more cloud-aware, so that the code tools now follow after is no big surprise but follows a greater trend. For cloud computing, it is another piece that falls in place, as cloud computing increasingly becomes the dominant way to run applications. Software-as-a-Service (SaaS) applications, as the new Visual Studio Online actually is one example of, get a bigger and bigger place in the application market.
This is an old idea in a new wrapping. Amazon (among others) now offers what they call Amazon WorkSpaces, which is basically a workstation running on Amazon WebServices. This gives users a Windows 7 environment with up to 100 GB local storage and 7.5GB memory. The price is not too bad either, between 35 and 75 USD per month.
The idea of having a remote workstation hosted on a bigger machine is very old, from the 1960s. Then a large mainframe was used to give users a share of the mainframe, and the local device just displayed the picture (or characters, this was in the 60s...) sent from the mainframe. Thin clients like this has since existed in various forms, and Amazon WorkSpaces is the latest incarnation of this idea. With services like Amazon WorkSpaces, there's no need to buy a large machine to host all the remote workstations on anymore, so the investments for starting with remote workstations is now substantially lower than earlier.
Analyzing large amounts of data in order to find trends and patterns requires vast amounts of computing power. Before cloud computing, such tasks required massive investments in servers and data center infrastructure, and only very large companies and governments were in reality able to do it. With the cloud, we have a scalable pool of computing power available, where we can turn on and off computing power as needed. This drastically lowers the cost of data analytics.
The effect of this is that small- and midsize companies can now perform data analytics and experience some of the benefits of this, which was previously reserved only for the big players. Detecting sales trends and which marketing campaigns are really working is now a matter of getting started with data analytics, and not figuring out whether it is possible. This is a major shift in thinking, and opens up options where even the smallest companies can operate cleverly and data-driven in their market.
Many of us have not been thinking about this, but cloud computing on a per-company scale has actually been here for quite some time. Both the massive web search engines like Google and Bing, and web shops like Amazon.com were running on the predecessor of today’s generic cloud computing platforms. And they evolved to what we see today, where cloud computing is available for rent from massive data centers, from the exact same companies which were behind the first truly global applications. The nature of these applications, with massive amounts of data to process and users all over the world, led to the development of large data centers and the beginning of the cloud computing industry. New players have also appeared on the cloud computing scene of course, but the global applications were the pioneers.
Cloud computing is becoming more and more similar to services like electrical power and water: They are delivered as needed and at a (mostly) reasonable price. And like electricity was when it was commercially introduced well over 130 years ago, cloud computing power will be (and is already) revolutionizing parts of our lives. And there's more to come!