Background Image

Blog Post

Nov 15

Parallel Computing Vs Cloud


Imagine for a moment you keying up a serious computer with real muscle. One of those fancy $40,000 Puget Systems jobs. You login to a server and see a distributed network waiting for a command, everything is pristine, a controller GUI - no even better a CLI list in front of you shows which resources on which machine are just waiting for your command. Now of course the question becomes, "Which process on which machine do you actually need to do what?" If what you needed was to solve a massive resource issue, you might need to break it into components and let the machine solve it in Parallels. 

Clustered Nodes

Parallel computing is mostly this kind of computing, 200 machines being controlled from one terminal. You don't need a $40,000 computer to do anything like that. You just need an interface that has the ability to issue commands across the networks of clustered machines and spit out solutions in assembly ready chunks. Not much practical purpose to that sort of thing at home, but maybe at work.

Clustered Nodes As A Cloud

Similar scenario, except a different problem altogether. You key into your $40,000 Puget, login and are faced with a readout of available resources slipping into and out of a variety of systems already in use. You notice that a central storage area has access to hundreds of Petabytes of space which isn't filling up very rapidly. You decide to try to solve the same resource issue and the machine files it in the cue for the next go around with processing. 

What Differences Do You Spot?

The number of users may seem obvious, but multiple users can make use of Parallels on distributed networks. The resource allotment cue is different, the process of getting at storage is perhaps somewhat different, and most certainly the functionality is different. You can use a cloud system as though it were a massive computer, multiple users can access the same files, generally at the same time. That means that if someone in Bolivia is watching the Matrix in their native language, you can still watch it in yours at the same time. What parallel computing does is more akin to staying on task with as much power as possible, where clouds generally distribute power in calculated shares on a per user basis.

Can The Same Hardware Do Both?

Yes - at a cost. There are configurations of cluster systems possible that make use of virtualization in ways that enable a wider assortment of user experiences and in some cases can perhaps integrate them well enough on a user basis somewhat smaller than a system dedicated to just one configuration or another. VDI's on distributed networks is a system as a service and in fact can be utilized for cloud desktop systems. Some systems exist which offer a degree of security that can be incorporated into your infrastructure without the same hassle as some of the other VDI's on offer. 

What Can This Do For Your Business

It can enable your business to make use of security technology in more compliant ways. A dog groomer might not need this technology, but a successful one who has a franchise under their management almost certainly does. It may be possible to individually secure a network for everyone who has a professional IT team at their disposal, but even then it's rare. Compliance with security protocols across large systems is akin to locking the backdoor at night. A smart business does it, and a foolish business never does. If you are interested in discussing this technology I can recommend you contact . Tell them "Always In Tao" sent you, to get a free consultation and ask how it could help your organization.

Back   
 
Add Comment:
Please login or register to add your comment or get notified when a comment is added.
1 person will be notified when a comment is added.