How Computer Chips Are Produced

This guide covers how computer chips are produced with practical context and easy-to-follow details. Use it to understand the subject and apply the information confidently.

• Chip design: This is the step that architects design chips, which means how it will work.
• Wafer fabrication: This is the main process in chip manufacturing and we will look into it in this tutorial.
• Prepare stereotypes: This step basically involves cutting chips from wafer
• Package: In this step, the terminal and the main part are added to the chip
• Test: The chip is tested before being sold.

Figure 1: The ingot is cut thin to create coarse wafer

Figure 2: How printing is done

Obviously each mask has different patterns and they are the way transistors and wires inside the chip are built. The number of masks used also varies depending on the given project. For example, Pentium 4 processors use up to 26 masks. Let's look at exactly how this process is done. The first thing that is done to the raw wafer is to implant silicon dioxide (SiO2) on it, by doing this by exposing the wafer to extremely hot temperatures and gas. This implant is similar to the way that iron adheres to a metal surface when exposed to a wafer, but it will happen faster. Next the wafer is made of a material called photoresist, which is a substance that can be dissolved when exposed to radiation. The first mask is applied and the wafer is exposed to radiation. The software of photoresist is separated by a solution and then the components of the exposed silicon dioxide layer are separated during the acid separation process. The excess of the photoresist is removed, so now we have a wafer with a silicon dioxide layer shaped like the first mask. Another silicon dioxide layer continues to be made on the wafer, large polysilicon is applied on its top surface and then another photoresist layer is applied on them. The second mask is made and the wafer is exposed to a second radiation. The soft photoresist part is removed with a solution and then the polysilicon parts and the silicon dioxide layer are exposed and separated by acid. The excess of the photoresist is removed and now we have a wafer with a silicon dioxide layer with a mold like the first mask and its upper part is a polysilicon and silicon dioxide layers have a pattern of the second mask. After these two steps, a process called doping (or ionization) will take place. Here the exposed areas of the wafer are bolded with different symbols, aiming to change the pathways of exposed areas to electricity. Exposed areas will be converted into P-type semiconductors (anodes) or S-type semiconductors (cathodes), depending on the chemicals used: Phosphorus, Atimom and arsenic are substances Typically still used to create N semiconductor layers, and Bo, Indium and Gallium are used to create P semiconductor layers. The arrangement of these semiconductor layers creates PNP or NPN transistors. Creating layers and using masks is repeated many times after the layout of the next mask. A metal will then be pressed into the wafer, filling the holes that have been formed to create a connection between the layers. The process of using masks and acid separation will be done to add electrical connections. This process is repeated many times until the chip is done, meaning the masks are used up. The production process and the number of layers depend on the components produced. For example, with Pentium 4 processors, they used 26 masks and 7 metal layers.

Figure 3: Transistors built inside the chip and connect the metal between them.

Figure 4: Wafer in Pentium 4 processor after being manufactured

The chips on the wafer after being tested and wafer are sent to the next step in the production process, in this step the chips will be wafer cut, with their end components attached and packaged. They will then be tested, branded and sold. All the process described above happens inside a room called 'clean room'. You may have seen some pictures of people working in this room with special clothes called 'bunny suits'.

Figure 5: Rabbit costume in clean room

Because we are producing micro-transistors so just a small dirty action can damage the chip, you can see the examples in Figure 6.

Figure 6: A bit of dirty cleaning can also damage the chip