One CPU is made up of highly complex circuits with millions of transistors. But how does a processor is made? Whether mobile smartphone, classic notebook, or large Windows PC, the heart of any computing device is the processor; without it the screen remains black. But how does an Intel or AMD processor actually are made?
Computer as large as cabinet walls
50 years ago, computers were as large as big cabinets and their energy requirements were insane. On top of that, these monsters needed an elaborate cooling, which brought a deafening noise. Half a century later, fingernail-sized processors in present in smartphones and tablets and they allow nimble and completely silent processing of photos and videos; they make the use of Internet and Office applications possible.
Billions instead of thousands of transistors
And how does that work? On Intel’s first processor 4004, there were 2,300 transistors in 1971 with a structure width of comparably thick 10 microns and an area of 144 square millimeters. Intel’s current “Kaby Lake” generation, on the other hand, houses an area of only 122 square millimeters and up to 30 plates of almost two billion circuits in a structure width of only 14 nanometers; 1 nanometer is 20,000 times thinner than a human hair.
Extremely high development costs
What has not changed is the raw material of processor, which is basically sand, to be precise, quartz sand (silicon.) Why? Quartz sand has high silicon content required for processor production. A finished CPU is actually created through a very complex process. The development of new processors consumes billions of dollars. In particular, when a radical change of generation is pending, in which manufacturers reduce the structural widths in order to accommodate an even greater number of circuits on the CPUs for a higher performance. For example, at the beginning of 2018, Intel wants to launch the first compute units with 10-nanometer structures, which are named “Cannon Lake.” Ice Lake processors are expected in early 2019.