The concept that most users have of a processor is that it is a simple piece of hardware that is placed on the motherboard and generates a lot of heat. However, a processor is made up of thousands of complex elements, which allows it to perform the mathematical operations necessary to make everything work, because ultimately everything the computer does must necessarily go through the processor, including what is processed by the GPU, which is why it is so important.
How a CPU is made
Although the operation of processors may seem magical, it is the result of decades of intelligent engineering. As transistors, the elements that make up most processors are reduced to microscopic scales, the way processors are made is increasingly complicated.
Photolithography is what brings processors to life
We are used to seeing wafers filled with dozens of chips which are then used in processors, but to get there you have to go through a series of steps that start with the photolithography.
Transistors are now so incredibly small that manufacturers cannot build them with normal methods. While precision lathes and even 3D printers can make incredibly complex creations, they typically achieve micrometric levels of precision (around thirty thousandths of an inch), but they are still not suited to the nano scales they are made of today. hui.
Photolithography solves this problem by eliminating the need to move complex machines with great precision. Instead, it uses light to burn an image onto a silicon chip, as if it were an old overhead projector that could be found in classrooms, but upside down, shrinking scale the model to the desired precision.
Thus, the image is projected onto a silicon wafer machined with very high precision on special machines (the famous machines made by ASML) and under extremely narrow conditions, as any dust spot on the wafer could mean that it will be completely spoiled. . The wafer is coated with a material called a photoresist, which responds to and reacts to light, leaving an etch of the CPU that can be filled with copper or other materials to form the transistors. Then this process is repeated several times increasing the size of the CPU in the same way that a 3D printer accumulates layers of plastic.
The problems of photolithography at the nanoscale
It doesn’t matter if you can make the transistors smaller and smaller if the transistors can’t work, and nanoscale technology has a lot of physics issues because of its size. Transistors are supposed to stop the flow of electric ity when they’re turned off, but they get so small that sometimes electrons are able to pass through them. It’s called the quantum tunnel and it’s a huge problem for silicon engineers.
Defects are another problem; even photolithography has a limit of accuracy, it is somewhat analogous to a blurry image from the projector, which does not show such a clear image when enlarged or reduced. Silicon factories are currently trying to mitigate this effect by using EUV (extreme ultra violet light) technology, a wavelength much higher than humans can perceive, using lasers in a vacuum chamber. However, this problem will persist as the size continues to decrease.
Sometimes the faults can be mitigated with a process called binning: if the fault affects a processor core, that core is deactivated and the chip is sold as the bottom part. In fact, most of the processor lines use the same model, but their cores are disabled because they are defective and therefore are sold at a lower price as a low end product.
If the defect hits, for example, the bezel or other critical component, the chip will likely have to be scrapped, resulting in lower manufacturing performance and therefore higher prices. Current process nodes, such as 7 and even 10 nanometers, have higher throughputs than 5nm nodes and therefore the reverse is true, their price is lower.
Packaging, essential in the manufacturing process of a CPU
After the process of manufacturing a processor, once we have the chips ready, they need to be packaged for consumer use, and that’s more than just putting them in a box with styrofoam. When a processor is finished it is still useless unless it can be connected to the rest of the system, so the process of “packing” or “wrapping” refers to the method in which the delicate silicon matrix ( matrix) is attached to the PCB that most people think of as the CPU.
This process requires a lot of precision, but obviously not as much as the previous steps. The CPU matrix is mounted on a silicon board and the electrical connections run on all pins that contact the socket on the motherboard. Modern processors can have thousands of pins, like AMD Threadripper processors which have 4096 pins.
Since the processor produces a lot of heat and also needs to protect its integrity on the other side, a built-in heat sink or English IHS is mounted on top. This comes in contact with the die and transfers heat out of the die, which we then cool using a CPU cooler. For some enthusiasts, the thermal paste used to make this connection is not good enough, which makes them decide to do a process of removing the processor.
Once everything is assembled, it can now be packed in real boxes, ready to put on store shelves and mount on our computers. Now that you know how a processor is made and the complexity of its manufacture, it’s a wonder most modern processors only cost a few hundred dollars, right?