High Performance Computing: What It Is and Why You Should Care

Anyone who works in advanced science or technology has heard the term “high performance computing” or HPC. This otherwise confusing terminology refers to scaling up ordinary computing power to perform high-level operations.
These operations may require processing ability that is hundreds to thousands of times greater than regular computers can provide. A well-known HPC application is the supercomputer, which use enough energy as multiple residential homes and produce an equivalent amount of heat.

The basic technical definition of HPC is aggregating computational power and scaling it up to achieve much higher performance levels. A single supercomputer contains tens of thousands of microprocessors. They measure operations in terms of trillions of bytes of data per second. The current operational level that has been achieved is eight petaflops, or eight quadrillion calculations per second, which was demonstrated by the K Computer in 2011.

Inside A High Performance Computer

A supercomputer contains the same parts as a desktop or laptop computer. The difference is there are more of these parts, and they are arranged in different ways. An HPC is actually a cluster of computers networked together to act as one giant device. Linking so many processors together is what allows the individual computational abilities of each to be combined or aggregated. Creating this network of nodes, a technical term for a single cluster, enables the computer to solve problems that the disaggregated computers cannot solve on their own.

The heart of a supercomputer is each processor. In the early days of computing, only one processor could be fit inside each machine. Today, the miniaturization process has shrunk common components to such an extent that engineers can fit two, four or even eight processors inside the space previously occupied by one. Intel Corporation's line of Xeon processor chips are known as quadcore chips because each chip has four cores or processors linked together. Supercomputers have thousands of cores to crunch unfathomably large numbers and perform inhumanly complex tasks.

Why HPC Is Important

High performance computing is stretching the boundaries of what was previously thought possible. Supercomputers are used in nuclear fusion research, underground oil exploration, cancer research, DNA sequencing and weather modeling to name a few applications. As a species, our ignorance of some of the most basic things is still vast. For instance, no one knows why tornadoes form or how cancer begins. Nuclear fusion, the dream of virtually limitless energy, can become a reality through continued research and supercomputing capacity. Though this is all based on large scale, we can also take HPC and see how it works for your business today. To learn more about those options and HPC in business visit www.dell.com.


The Future

Frontiers for HPC even still exist. Despite the many milestones that have been reached and overcome through the years, the dream of exascale supercomputers pulls research forward. This will require designing, building and testing supercomputers a thousand times more powerful than eight petaflops.

Intel Corporation opened three research centers dedicated to exascale supercomputing last year. The Defense Advanced Research Projects Agency or DARPA as well as private industry are laying the groundwork for this mind-boggling advance in computing power. Exascale research may yield results before the end of this decade, and possibly even sooner.

Author Bio

David Malmborg works with Dell. When he isn’t working he enjoys hiking, spending time with family and researching new technology. He is currently learning more about enterprise solutions and recommends visiting the Dell website to learn more about solutions for your business.


To write guest posts for us, please follow the link below

Write Guest Articles for us

No comments:

Post a Comment