Have you ever wondered how a processor is made ? They are the most complicated and complete devices in the world, and the difficulty to improve them is increasing. Now Intel has made a post where they explain to us how they make their processors in a way that we can easily understand.
Processors are not really limited to just PCs or servers, but actually almost all devices use processors of some kind: from your alarm clock to your smartphone. However, and as the one who published the video is Intel, in this case we are going to explain how a brand processor is manufactured or what it consists of, starting with how it is designed to the final product, which is the one that users install and use in our personal computers.
How an Intel Processor Is Made
In a space not much larger than a fingernail, a processor contains billions of microscopic “switches” called transistors , which are what make a processor work.
Once we know that, we are going to see the manufacturing process of the processors, from their design until they reach our homes.
The design of the processors
Before becoming the “square” that we receive in our house when we buy a processor, everything begins with a simple idea in the head of the company’s architects. These architects work in conjunction with engineers and designers to create the initial sketches of how the device is to function. These sketches, when all agree, become the final drawings , filled with transistors, circuits and layers.
These layers are important, and a processor can have more or less 30 of them. Some layers contain the transistors, and others the interconnections between the different layers in specific configurations to maximize their efficiency.
Transistors, like “switches” that they are, can be turned on and off to represent the ones and zeros of the binary system used in computing up to 5 billion times per second.
Creating the template and mold
Once designers, engineers and architects are happy with the design they have captured in the plan, this design is sent to the “Mask Ops”, engineers who are in charge of translating the design into a template that can then be used to manufacture the processor.
To do this, an Electron Bean Machine replicates this design in 6 × 6-inch pieces of quartz , 1/4 inch thick. These pieces are called masks (Mask), and they are the ones that are used later to shape the internal circuitry of the processor in a silicon wafer . A kind of mold, and it takes more than 50 Masks to be able to make all the layers of the processor.
The manufacturing process
Once all the necessary masks to manufacture a processor have been created, it is passed to the manufacturing phase and these Masks are sent to the factories, known as Fabs. This is where these molds will be used to shape the circuits on the wafers, which you have surely seen many times before. Obviously, these wafers do not exist in that state, and first they have to go through chemical processes to convert the sand (silicon) into the wafers that we know.
To “print” the masks on the wafers, a process called photolithography is used, whereby an electron gun reflects the light on these masks, which by passing through different lenses to reduce the process to the necessary size, they are primed in wafers.
This must be done with all the masks to create the layers of each chip. Thus, hundreds or even thousands of small chips can fit on a single wafer. And, once done, you move on to the next step in the process of how a processor is made.
The preparation process
Once we have the wafers, we move on to the preparation and ordering step. Basically, a wafer contains hundreds or thousands of chips, and you have to cut them with total precision to be able to separate all these chips from each other so that we can later use them in the processors. For this, extremely precise laser cutting machines are used.
The result of this process is the die we all know, which is the brain of processors. Once the chips are cut, another machine transports them to the next phase of the preparation chain.
In this next phase, the chips are put in a kind of rolled-up tape so that they can travel by plane, since they have to be sent to other Intel Fabs: those for assembly and testing.
Assembly and testing
In this phase, engineers test all chips individually and discard those that do not work properly or those that do not pass the manufacturer’s quality standards. If they pass the test, the chips are mounted on a substrate and topped with the heat sink (IHS) , creating what we all know as a processor. This process is called assembly.
This outer packaging protects the chip from almost all damage, including bumps, splashes, or heat. This substrate has in its lower area all the contacts necessary for the processor to work in conjunction with the motherboard where we install it, of course.
Once the assembly process is finished, the processor already goes to the last step before it reaches us, the stored one.
Inventory and warehouse
In this last step, the processors are put into their boxes, along with the heatsinks, instruction manual and so on, and they are all packed together. This is the product that we will buy in the end.
From here, Intel sends its processors to OEM manufacturers, distributors and the entire sales network it has around the world, who will either sell the PCs with a processor already mounted, or they will serve the material to the stores that is where we are. we can buy a processor.
You have already seen it. From the moment the processor is designed until it reaches our homes, it goes through a complex, time-consuming process that takes place around most of the world and involves hundreds of people.
