It used to be common in the computer industry not long ago to require separate PCI video cards to enable video processing, as the processing chips didn't include that feature among their limited portfolios. These cards were the predecessors of what we now called Graphics Processing Units (GPUs) which differentiate themselves from the pure CPUs in that they work more efficiently processing complex arithmetic algorithms in the mostly work on the Of course, back then the majority of basic software applications in use didn't require much processing power in general. However as the silicon processing technology has progressed, ever more peripherals are included in the logic cores of the microprocesor. One of which is the graphics processing unit
Specialty niches were emerging thouth, to satisfy the particular needs of selected tech savvy users who desired much more processing ability from their machines; so specialty design houses sprung to satisfy those needs and benefit commercially. One among many was NVIDIA
The ever decreasing scale in microelectronics
Ever since the beginning of the microelectronics era there has been an eternal quest to reduce the characteristic features on the devices: some devices are now in qualification states on the sub 40nm gate oxide range for an scheduled commercial release towards the end of the year, and there are a lot of efforts in the sub 30nm range. But Moore’s Law, as this continuous drive to reduce sizes has come to be called, applies not only to the dimensions on the silicon. It spreads down through the whole microelectronic supply chain, into pcb packages, system integration and final devices.
br>
br>
br>
br>
br>
|