Have you noticed the latest white paper advertised on your favorite tech news site? Or maybe you have noticed it in the news itself from your neighborhood technology company. Cloud Computing is the future baby! If you did not know it yet, please try harder to get out from under that rock.
Make no mistake, I do love the technology and the possibilities opened up by what is called as Cloud Computing. Moving the infrastructure Shipping and Handling out of individual responsibility and aggregating and delivering it remotely reduces a lot of pain (most of the time) so you can focus on adding value at the next layer. It is the next step in the march from moving computation boxes from under the desk to the data center. If your customers and employees are going to the information you supply over the web, they do not need to know or care whether it coming from under your legs, from your building or from AWS. However, the hype around the Cloud Computing term is going into excessive overdrive.
Wikipedia article on Cloud Computing has a paradigm shift in the first sentence for the description. So it is not just the marketing folks at technology (and shampoo companies with cloud ambitions) who are frothing at their mouths. Everyone seems to be caught up in the frenzy.
I would recommend everyone (okay maybe only the techies) to browse through the following classic paper from 1968 which I read occasionally to remind myself about concepts that re-invent the wheel:
On the Design of Display Processors (Myers, Sutherland)
Yeah, I know you gave up after a few paragraphs even if you opened that link. Do give it a try at leisure though.
Technology makes progress at a frenetic pace. Transistors go faster and smaller, cables become fatter, signals transmit faster. Every once in a while you realize that you can actually adapt concepts you used to apply to large scale systems and map them to small scale systems and vice versa. The individual processing systems that used to be large can now be compressed into small blocks within a chip. You can apply lessons learned at connecting systems and machines together using networks and apply them to connecting blocks within a chip. The field of interconnection networks was hot late in the 1990’s applying knowledge from high speed switching and routing within mainframe systems back into interconnection networks within multi-processors and network routers.
Imagine being hunched over a vt100 dumb terminal blinking green and working away at tasks that actually were running on a mainframe computer a few walls away in an air-conditioned locked room. Or maybe you actually needed to be in that air conditioned room. Now make that connection from the terminal much longer and going under a few miles of dirt and few thousand miles of ocean floor. However, make the latency almost same. And make that interface protocol a wee bit more complicated than text terminal display (HTTP, BigTable, MapReduce, Hadoop, Hive, Pixie Dust). Increase the marketing budget by a few million dollars. You get the new Cloud Computing Wheel.
Here is that brand new wheel for you. That rock you were under must have mangled your gray cells to call it the same as the original one:
|
|
