Cloud Parallel Processing

Cloud Parallel Processing

The current state of Cloud Parallel Processing and how it might affect the individual user

Parallel processing is essentially multithreading on a much larger scale. Multithreading is utilized in situations involving both hardware and software; its purpose is to essentially 'divide and conquer', or evenly distribute workloads so that functions can be accomplished faster, powerfully and more efficiently. In most instances, multithreading is employed as a hardware function; traditionally, software multithreading is not used (at the individual level) to process tasks that place a heavy burden on system resources.

In cloud computing parallel processing might entail using multiple versions of the same operating system, program, virtualized hardware or simultaneous versions of the same machine in order to boost computing power and speed. The problem with parallel processing (as it relates to cloud computing) is that there is currently a lack of language or code to adequately support, integrate or utilize it.

The obvious solution (for lack of resources within the context of cloud computing) is to simply give individual users the capability to requisition the additional resources needed to perform a task or series of tasks. The problem with this however is that many (nearly all) of the applications / programs that run within the cloud computing framework are not currently fully compatible with parallel processing. Or perhaps a better way of putting it would be; they are not capable of taking advantage of everything that parallel processing has to offer. This of course means that software designers (as well as cloud developers) are going to need to find or create new ways of adapting these applications or programs for use in a multi-core/processing situation.

Basically there is a huge potential here for individuals, businesses and cloud computing in general. The power and scope of parallel processing is vast and untapped, once more developers become aware of the inherent potential (of parallel processing) there will be a surge in this area. A comparable illustration of the current scenario; it would be like giving an automobile to a rural town that's only familiar with the horse and cart, it will take them a while to figure out how to drive this new invention as well as how to best use it. The most obvious place to begin employing parallel processing is within the realms of multimedia. Parallel processing could also put an end to many of the online performance issues that consumers are apt to complain about. Despite the prevalence of broadband services worldwide, many individual users are still somewhat hampered by the slow performance of online video, gaming, as well as other types of multimedia viewing.

Just imagine being able to run multiple versions of the same program without error or performance lag. For example, programs like Photoshop (which is known to over exploit and hog system resources) utilize multiple threads to perform their operations. If a graphic designer wants to work on multiple projects simultaneously they have to open and run them within one application framework. Not only does this bog down their system, but it also can limit them creatively. With parallel processing they could run entirely separate versions of the same program all at the same time without any performance lag or deterioration. This could eventually lead to a re-thinking of the way that software is designed and implemented, paving the way for new programs that are set up to run multiple instances instead of just threads.

There are essentially 2 schools of thought when it comes to the current state of cloud computing. One sees it as the phasing out of the era of the personal computer; with its own hardware and software. The other sees it as a bridge to a new way computing on a grand scale. Both of these observations are true in a sense, but where one group may see something as being wholly negative, the other sees potential. Obviously there will be growing pains; the current model of computing didn't simply spring up overnight either, it took many years of toil, trial and error to reach its current state. The real argument in the 'for or against cloud computing' debate is really about how all these past discoveries are being utilized in a future model. How does this relate to parallel processing? In many ways parallel processing is one of the next evolutionary phases in the development of cloud computing itself. The power that parallel processing offers to the individual is unsurpassed and unprecedented in the history of personal computing. For the average person to attain the kind of power offered by PP they would have to invest in a very expensive system, which would be far outside of the means of the majority of individuals. In truth however, the power offered to the individual by a parallel processing system is virtually limitless; accruing that same level of power would essentially entail the individual user building their own private cloud, which is extremely unreasonable. In the end it all comes down to whether or not you would like your computing service(s) to be akin to a utility (water / electricity) or something that you self manage.

Once the kinks have been worked out and parallel processing becomes commonplace (assuming that they are able to work out the inherent problems) the individual is going to be able to do more than ever before, and all for a very slim price tag. On a grander scale, parallel processing might very well raise global expectations for the speed and efficiency of delivered data over the net. The prevalence of parallel processing could very well lead to a fundamental shift in consumer expectations as well.

+++

Would you like to learn more about Cloud Computing and how to manage your IT Service through Cloud technology? Sign up for our Cloud Computing Foundation Program. Now also available for iPad and iPhone.