Graphic Processing Units

572 Words Jan 28th, 2018 2 Pages
A GPU structure is built in a fashion so that it can process parallel operations on large block of data. This feature of parallel processing of data give an advantage to GPU over general purpose CPU’s for algorithms processing. NVidia is the first company who has the credit to launch first ever GPU (GeForce 256)1999.
GeForce consist of 22 million transistors as compare to 9 million transistors on Pentium III chip and GeForce has the capability to process 10 million polygons in a second. The speed and performance of GPU is directly proportionate to its sophisticated architecture. GPU has its own memory on stand-alone cards, where as in the chipset or CPU chip it share main memory with the CPU. Normally GPU’s are found on plug-in cards, in a chipset on the motherboard or in the same chip as the CPU.
Considering the utility and advantage of GPU it is now vastly use in mobile devices, Graphic cards of laptop, game consoles and many other electronic devices.
Initially GPU was designed for graphic purpose only. It is based on vector matric calculation. But a student from the Stanford University found that GPU can be used for general purpose. From then different software has been develop that support GPU, and programming become much easier for latest version of GPU compare to the initial version of GPU. And this…
Open Document