Silicon, which is extracted from sand, is melted in a furnace. A tiny seed crystal of silicon is added to the red-hot molten silicon. A big crystal grows around it and is slowly pulled out, forming a long, sausage-shaped crystal of pure silicon. The silicon sausage is then cut into very thin slices, called wafers. Silicon is used because its electrical properties can be changed by adding impurities.

Each wafer is heated to create a layer of silicon dioxide, which is then given a light-sensitive coating and exposed to ultraviolet light through a mask. The light hardens the coating in some places. In others, the oxide can be etched away, leaving a pattern of naked silicon ready for doping.

The silicon wafers are heated in a furnace full of a gas containing another element, such as arsenic. This process, called doping, adds impurities to the silicon, altering its electrical properties. Different combinations of heat and chemicals form transistors and other components on the silicon. Each wafer goes through many stages of masking, etching, and doping.

Computer-controlled testing equipment puts each wafer through a set of tests to make sure that every chip is working properly. Even though operators wear protective suits during the manufacture of the chips, some chips are spoiled by just a speck of dust. Failures are marked so that they can be recycled once the wafer has been cut into chips.

Metal pins are connected to the chip by welding fine gold wires to the pins and chip. The chip is then encased in a protective plastic or ceramic package, leaving the pins sticking out. The pins are then soldered into a thin plastic board with copper tracks “printed” on it. This connects several chips to form a circuit. Some chips, such as memory chips, are placed in sockets.