CPU, Central processing unit. It is the core component of modern computers, also known as "Microprocessor (Microprocessor)". For PCs, CPU specifications and frequency are often used as important indicators to measure the performance of a computer. The Intel x86 architecture has been around for more than 20 years, and the x86 architecture CPU has had a profound impact on the work and life of most of us.
Classic CPUs from generation to generation
Many friends who know a little bit about computers will know that the most important thing in the CPU is the transistor. Improving the speed of the CPU is the most important thing To put it bluntly, it is how to put more transistors into the same CPU area. Since the CPU is too small and too precise, it contains a considerable number of transistors, so it is absolutely impossible to do it manually. It can only be done through photolithography. Processed by technology. This is why there are so many transistors in a CPU. A transistor is actually a two-position switch: on and off. If you recall the days of basic computing, that was all a computer needed to do its job. The two options, on and off, are 0 and 1 for the machine. So how would you make a CPU? In today's article, we will tell you step by step the entire process of a central processing unit from a pile of sand to a powerful integrated circuit chip. (Due to the high technical content of the CPU production process, the editor's capabilities are limited, so the pictures and introductions are collected from the Internet). This article is just to give everyone a more detailed understanding of the CPU production process, so that the editor's task is completed.
● The basic raw material for manufacturing CPU
If you ask what the raw material of CPU is, everyone will easily give the answer - silicon. This is true, but where does silicon come from? In fact, it is the most inconspicuous sand. It's hard to imagine that the CPU, which is expensive, complex in structure, powerful in function, and full of mystery, comes from that worthless sand. Of course, this must go through a complicated manufacturing process. However, you can’t just grab a handful of sand to make raw materials. You must carefully select and extract the purest silicon raw materials. Just imagine, if you use the cheapest and abundant raw materials to make a CPU, what will the quality of the finished product be like? Can you still use such a high-performance processor as now?
Intel technicians use automated measurement tools in semiconductor manufacturing plants to monitor wafer manufacturing progress according to strict quality standards.
In addition to silicon, another important material required for manufacturing CPU is metal. So far, aluminum has become the main metal material for making internal components of the processor, while copper is gradually being phased out. There are some reasons for this. Under the current CPU operating voltage, the electromigration characteristics of aluminum are significantly better than copper. The so-called electromigration problem means that when a large number of electrons flow through a conductor, the atoms of the conductor material are impacted by the electrons and leave their original positions, leaving vacancies. Too many vacancies will cause the conductor connection to be disconnected, and the atoms that have left their original positions will Staying in other positions will cause short circuits in other places and affect the logic function of the chip, making the chip unusable. This is the reason why many Northwood Pentium 4s were replaced with SNDS (Northwood Burst Syndrome). When enthusiasts rushed to overclock the Northwood Pentium 4 for the first time and greatly increased the chip voltage, serious electromigration problems caused the CPU of paralysis. Such was Intel's first foray into copper interconnect technology, it clearly needed some improvements. But on the other hand, the application of copper interconnect technology can reduce the chip area. At the same time, because the resistance of copper conductors is lower, current can pass through them faster.
In addition to these two main materials, some types of chemical raw materials are also needed in the chip design process. They play different roles and will not be described here.
● Preparation stage of CPU manufacturing
After the collection of necessary raw materials is completed, some of these raw materials need to undergo some preprocessing work. As the most important raw material, the processing of silicon is crucial. First, the silicon raw material is chemically purified, a step that brings it to a grade that can be used by the semiconductor industry. In order to make these silicon raw materials meet the processing needs of integrated circuit manufacturing, they must also be shaped. This step is completed by melting the silicon raw materials and then injecting liquid silicon into a large high-temperature quartz container.
The squares on the wafer are called "die". Each microprocessor will become the "brain" of the personal computer system.
Then, the raw materials are melted at high temperature. We learned in middle school chemistry class that the atoms inside many solids have a crystal structure, and the same is true for silicon. In order to meet the requirements of high-performance processors, the entire silicon raw material must be highly pure and monocrystalline. Then the silicon raw material is taken out from the high-temperature container by rotating and stretching, and a cylindrical silicon ingot is produced. Judging from the process currently used, the diameter of the circular cross-section of the silicon ingot is 200 mm. But now Intel and some other companies have begun using 300 mm diameter silicon ingots.
It is quite difficult to increase the cross-sectional area while retaining the various characteristics of the silicon ingot, but it can still be achieved as long as companies are willing to invest a lot of money in research. Intel spent approximately $3.5 billion building a factory to develop and produce 300mm silicon ingots. The success of the new technology allows Intel to manufacture more complex and powerful integrated circuit chips. The 200mm silicon ingot factory also cost US$1.5 billion. Let’s start with the slicing of silicon ingots to introduce the manufacturing process of the CPU.
Clean air flows continuously into the room from the gaps in the ceiling and floor. All the air in a clean room is replaced several times a minute.
After making the silicon ingot and ensuring that it is an absolute cylinder, the next step is to slice the cylindrical silicon ingot. The thinner the slice, the less material is used, and the processor can naturally be produced. There are more chips. The slices are also mirror-finished to ensure the surface is absolutely smooth, and then checked for distortion or other problems. The quality inspection at this step is particularly important, as it directly determines the quality of the finished CPU.
Some substances must be mixed into the new slice to make it a real semiconductor material, and then transistor circuits representing various logical functions are carved on it. The atoms of the incorporated material enter the gaps between the silicon atoms, and atomic forces act on each other, giving the silicon raw material the characteristics of a semiconductor. Today's semiconductor manufacturing mostly chooses CMOS process (complementary metal oxide semiconductor). The word complementary refers to the interaction between N-type MOS transistors and P-type MOS transistors in semiconductors. N and P represent the negative electrode and positive electrode respectively in electronic technology. In most cases, the slices are doped with chemicals to form a P-type substrate, and the logic circuits scribed on it are designed to follow the characteristics of nMOS circuits. This type of transistor has higher space utilization and is more energy-saving. At the same time, in most cases, the emergence of pMOS transistors must be limited as much as possible, because in the later stages of the manufacturing process, N-type materials need to be implanted into the P-type substrate, and this process will lead to the formation of pMOS transistors.
After the incorporation of chemicals is complete, standard slicing is completed. Each slice is then heated in a high-temperature furnace, and a layer of silica film is formed on the surface of the slice by controlling the heating time. By closely monitoring temperature, air composition and heating time, the thickness of the silica layer can be controlled. In Intel's 90nm manufacturing process, the width of the gate oxide is as small as an astonishing 5 atoms thick. This layer of gate circuit is also part of the transistor gate circuit. The function of the transistor gate circuit is to control the flow of electrons within it. By controlling the gate voltage, the flow of electrons is strictly controlled, regardless of the size of the input and output port voltages.
The final step in preparation is to cover the silicon dioxide layer with a photosensitive layer. This layer of material is used for other control applications in the same layer. This layer is very photosensitive when dry and can be chemically dissolved and removed after the photolithography process.
● Photolithography
This is a very complex step in the current CPU manufacturing process. Why do you say this? The photolithography process uses light of a certain wavelength to carve corresponding grooves in the photosensitive layer, thereby changing the chemical properties of the material there. This technology has extremely strict requirements on the wavelength of light used, requiring the use of short-wavelength ultraviolet rays and lenses with large curvatures. The etching process can also be affected by stains on the wafer. Each etching step is a complex and delicate process. The amount of data required for each step of the design process can be measured in units of 10GB, and more than 20 etching steps are required to manufacture each processor (one layer of etching for each step). Moreover, if the etched drawings of each layer are enlarged many times, they can be compared with the map of the entire New York City plus the suburbs, and are even more complicated. Imagine shrinking the entire New York map to an actual area of ??only 100 square millimeters. On the chip, you can imagine how complex the structure of this chip is.
Single crystal silicon ingot and initial core structure
When all these etching tasks are completed, the wafer is turned over. Short-wavelength light shines on the photosensitive layer of the wafer through the hollow notches on the quartz template, and then the light and template are removed. The exposed photosensitive layer material is removed chemically, and silicon dioxide is immediately generated under the empty position.
Intel technicians monitor wafers in an automated wet etch tool, a process that removes excess operating additives or contaminants from the wafers.
● Doping
After the remaining photosensitive layer material is removed, what remains is the silicon dioxide layer filled with trenches and the exposed silicon layer below the layer. . After this step, another silicon dioxide layer is created. Then, another polysilicon layer with a photosensitive layer is added. Polysilicon is another type of gate circuit. Because of the metal raw materials used here (hence the name metal oxide semiconductor), polysilicon allows the gate circuit to be built before the voltage at the transistor array port is applied. The photosensitive layer is also etched by short-wavelength light through the mask. After another step of etching, all the required circuits have been basically formed.
The exposed silicon layer is then chemically bombarded with ions, where the purpose is to create an N-channel or P-channel. This doping process creates all the transistors and their circuit connections. Each transistor has an input and an output, and the connection between the two ends is called a port.
● Repeat the process
From this step you will continue to add layers, adding a silicon dioxide layer and then photolithography once. Repeat these steps, and a multi-layered three-dimensional architecture emerges, which is the embryonic state of the processor you are currently using. Metal coating technology is used between each layer to conduct conductive connections between layers. Today's P4 processor uses 7 layers of metal connections, while Athlon64 uses 9 layers. The number of layers used depends on the original layout design and does not directly represent the performance difference of the final product.
● Testing the packaging and testing process
In the next few weeks, the wafer needs to be tested step by step, including testing the electrical characteristics of the wafer to see if there is any Logical errors, if any, at which level they occur, etc. Then, each problematic chip unit on the wafer will be tested individually to determine whether the chip has special processing needs.
Technicians are inspecting each wafer to ensure that each wafer is in optimal condition. Each wafer may contain hundreds of chips.
Close-up of the wafer rotating during testing
Then, the entire wafer is cut into individual processor chip units. Units that fail during initial testing will be discarded. These cut chip units will be packaged in a certain way so that they can be smoothly inserted into a motherboard with a certain interface specification. Most Intel and AMD processors are covered with a thermal layer. After the finished processor is completed, a full range of chip function testing must be carried out. This part will produce different levels of products. Some chips have relatively high operating frequencies, so they are marked with the names and numbers of high-frequency products, while those chips with relatively low operating frequencies are modified and marked with other low-frequency models. These are processors with different market positioning. And some processors may have some shortcomings in chip functionality. For example, if it has a flaw in the cache function (this flaw is enough to cause the vast majority of CPUs to paralyze), then they will be blocked from some cache capacity, reducing performance, and of course lowering the price of the product. This is Celeron And the origin of Sempron.
After the CPU packaging process is completed, many products need to be tested again to ensure that there are no omissions in the previous production process and that the product fully complies with the specifications without deviation.
The above is the entire production process of the CPU. I believe everyone will enjoy it, right? For those who want to know more about CPU, this article can meet your needs.
But for a deeper understanding of the more detailed principles of CPU production, you still need to find some professional information to study. I will not explain them one by one here. I hope you can understand