With penetrating depths of Internet, there is ever increasing demand to understand and implement supra-computing applications. Not just that, but applications in science, e.g. origin of universe, require computation capacity that just can not be met by commercially available hardware (even a gigantic combination of them). Let us take a tour de modern computing to peep in to future of networking and problem solving. Starting from the simple, home computing, a giant step forward is the grid computing and parallel computing. The cloud computing has also found its niche in web-based computational challenges, despite of the usual face-off with Internet speed limits.
Forerunner in the field of commercial computing, are Intel, who have unveiled plan to launch new CPU, i7 (code name: Nehalem). It boasts 3.2 Ghz speed, 4, 6 or even up to 8 CPU cores (731 Million Transistors for 4 core), space for three memory channels (L3 cache), and integrated graphics card. That’s not all, for the first time in the history, i7 would be implementing simultaneous multi-threading (SMT), miniature parallel computing per core. Intel calls it hyper-threading which improves the overall efficiency by processing multiple threads (2 or 3) per core. Sources reveal that i7 will boost performance by 15-20% with just 10% increased power compared to (existing) dual core processors.
One of the emerging field after the advent of internet high-ways, is the “distributed” or “grid” computing that utilizes commercially available hardware (processors, memory, etc.) connected to a network. The grid computing proves to be cheaper alternative to building a super-computer of equal capacity with dedicated local bus for data-transfer. It is ideal solution for parallel computation of several independent tasks. The most successful implementation for distributed computing is folding@home (Stanford University) which exceeded 4 petaFLOPS (4×10^15) milestone. The main usage is to solve complex protein phenomena and related diseases, basically via molecular dynamic simulation. Trivially, this performance was enabled by array of PlayStation 3!
Image: http://www.estilofacil.com
The third high-speed variant in the computing application is the “parallel” computing. Traditionally, all computing tasks are broken in pieces and solved “serially” at very high speed. The main components involved are the single central processor and the local data-transfer bus. In parallel computing, the task is broken in discrete parts that can be solved concurrently and involves multiple resources such as processors. Again, the processors involved can be completely local, sitting on dedicated network, or combination of both. Currently parallel computing is used to understand galactic formations, traffic predictions, aerospace engineering, and bunch of others. In many ways, parallel computing is similar to grid computing in that if the resources belong to shared platform. One limitation parallel computing puts on the problem itself is that, the problem must have independent executable blocks, nevertheless it has been applied to several critical time consuming tasks world wide.
Image: https://computing.llnl.gov
The last one, and deceivingly the most popular, is the cloud computing. This type of computing relies heavily on Internet for creating and sharing virtual resources, typically storage systems. It provides common business and personal applications through web browser. Yahoo’s Flickr for photos, delicious for tagging websites and other email and office applications, they all form the “cloud”. In fact, it has turned out to be the most economic, scalable and agile computing method so far! The customers do not own the resources that are part of the service and pay only a fraction of the total cost of it, depending on the usage. The architecture is fairly simple and secure, ranging from public, private to hybrid. The most important advantage of cloud computing is that, it is device and geography independent!
Image: http://www.wikipedia.org
So much so far! I hope to bring back more in-depth analysis in forthcoming articles, till then, Chao!
Read more on the same topic:
@Welcome gundappa
Hi Amol
mine blog also got PR2
Hello Amol,
)
Good article. Perhaps you could also throw some more light on application of the various computing technologies. I am currently involved in risk assessment of a project related to cloud computing. I could see the main application of cloud computing in mobile technology. Cloud computing would be part of the next generation mobile technology in terms of empowering them with data storage and processing capacity. God Bless Internet!
@Atul: Good to know that, buddy!
@Kedar: Great! It is indeed true that cloud computing is the next step in Internet, and with rapidly available openware it will bring down costs further. Android, the applicationware from google is one such example, it offers personalization!
I will definitely come up with applications, excellent idea.
[...] : Amol Naik What : Computing Universe Spicy : A Tech post which talks about the recent trend known as Cloud Computing. For more on it, [...]
Hello Amol
Congratulations! Your post has been selected by BlogAdda as one of the top posts for this week’s ‘Spicy Saturday Picks’.
Do mail me so that I can send you over something. It was a very well researched post.