High Performance Computing At Its Best

By Author: Ivan Cuxeva Jr
Total Articles: 779
Comment this article

Computers have undoubtedly changed the world, is because of them that communications are extremely fast, science has advanced so much and we are able to do things we never though even to be possible before, as time goes by new technologies emerge and current applications are improved in order to maximize their efficiency, this this case simple processors can be adapted to work in a high performance computing environment which can be used for in a number of ways, for instance high performance computing (HPC) can be used to explore the space, analyze atomic particles, render complex animations, process millions of calculations per second or even handle multiple business transactions.

Anyway you see it, the right mixture of machine code and hardware has helped us accomplish a lot. The phrase "high performance computing" does not refer to home PCs which are fast, this term is used to describe several clusters of microprocessors put together in order to handle extremely complex processes, the core of such machines is not especially built for complex applications, in fact these high performance computers are made of the ...
... very same processors available to the everyday consumer, however in order to build what is known as a HPC it is necessary to acquire these parts in bulk and then assemble them into clusters which will become part of the super computer.

At this point you might be thinking of a mainframe computer however there is a big difference since mainframes are for the most part monolithic in nature compared to HPC. To better understanding of what is considered a HPC we must establish a performance measure, in this case we will refer to the term FLOPs which stands for "Floating Point Operations per Second", any computer system which operates at a Teraflop can be considered a high performance computer; currently IBM and NEC have developed HPCs which operate at incredible speeds for instance, IBM's Blue Gene/P operates at one PetaFlop but when configured differently it can reach 3 PetaFlops; NEC's system on the other hand has a performance of 839 TeraFlops and features the first Vector processor which is said to be able to exceed 100 GigaFlops per single core; now that's heavy duty!

The world of super computers and high performance ones is just starting to emerge, we are sure to hear about record breaking systems within the next few years.