The upgrade of the chip -- optoelectronic chip, light instead of electricity?
Contemporary people are not only increasingly inseparable from all kinds of intelligent electronic products, but also have increasingly high requirements for the performance of electronic products. They always hope that the reaction speed of computers and mobile phones will be faster.
However, both smartphones and computers cannot be separated from chips, and the quality of chips determines the performance of electronic products. Today, electronic technology is changing rapidly, and a variety of new chips are springing up.
One very unusual class of chips uses light of different wavelengths instead of electric currents to transmit and process signals. These are called optoelectronic chips.
What's going on here? Below, let's introduce it to you
What is a chip?
Before we talk about optoelectronic chips, let's look at traditional chips. A chip's full name is integrated circuits, or, in general, circuits is a collection of many circuits with complex functions placed in a small area.
Take the most important CPU in a computer, the central processing unit (CPU), which is essentially made up of billions of "small switches" and other circuit components.
This little switch is called a transistor. When the transistor is on, an electric current passes through it; In the off state, the current is pinched off.
When the transistor "on", there is a current through, recorded as the state "1"; When the transistor is "off" and no current is passing through, it is denoted as state "0".
The chip in the computer, is by recording and processing hundreds of millions of "010101" number to achieve logical calculation and information processing function.
When the information is processed in the form of "010101", the numbers are translated into pictures, words and videos on the screen for us to watch.
So how exactly do these "little switches" in computer chips work?
Since current is required to pass through the inside of the transistor, there will naturally be an incoming and outgoing end of the transistor, called the source and drain, respectively.
The part that acts as a "switch" acts like a fence, blocking or allowing the current to pass through, so it is figuratively called the "grid".
Transistor diagram
Above is a schematic of a transistor, which is normally insulated so that no current can pass through its interior, i.e. in a "0" state.
At this point, if a voltage of a certain magnitude is applied to the gate, the conductivity of the transistor will change, forming a conductive channel, so that current can enter from the source, flow out from the gate, and the transistor will change to a "1" state.
The enabled chip
Thus, we can easily change the state of the chip to represent a "0" or a "1" simply by controlling the voltage of the transistor gate.
Hundreds of millions of these transistors, lined up together, can record and process huge amounts of data of all kinds, which can be translated into the picture we see.
Optoelectronic chip components
The principle of optoelectronic chips is actually very similar to that of traditional chips. They are also made up of numerous "small switches" (optoelectronic chip components). One technique is to replace the current in the "small switches" with light. Let's look at its structure first (simplified model) :
Schematic diagram of a simplified photoelectric chip element
It can be seen from the above simplified model that the optoelectronic chip components are mainly composed of grid, optical waveguide and substrate. The optical waveguide acts as a "sieve", allowing light of certain frequencies or wavelengths to pass through while blocking light of other frequencies.
However, the "sieve" of the optical waveguide is not static. Once a voltage is applied to the grid, it changes its screening criteria, allowing only a different frequency of light to pass through and blocking the rest.
The role of grid in optoelectronic chip components
Some of you may ask, why does gate voltage change the selection criteria of optical waveguides? How can optical waveguides screen for light of different frequencies?
In fact, this is similar to the principle of a guitar, a fixed string, when you pluck it always produces the same frequency of sound. But when you change the length and relaxation of the string by pressing with the other hand, the frequency of the sound will change.
Light is similar to sound in that it is essentially an oscillation of a wave, and as it enters the optical waveguide, it oscillates continuously inside. However, the properties of the optical waveguide boundary determine the mode and frequency of oscillation allowed in the optical waveguide. Only light with a specific frequency can oscillate in the optical waveguide.
If the frequency of light does not correspond to the optical waveguide, the light cannot oscillate in the optical waveguide, the oscillation disappears, the light will not exist, and the light that does not meet the frequency requirement will be blocked out.
The grid is like the left hand of a guitar, changing the nature of the boundary of the waveguide to change the frequency allowed to pass through it. Applying voltage is a simple and easy way to change boundary properties.
At this point, friends might as well think about yourself, if you use photoelectric chip components to make a CPU can be used to process, how would you design it?
In fact, it is very simple, we can follow the traditional chip design. First we choose a particular frequency of light that can pass through when a voltage is applied by the grid, denoted as state "1"; When the grid does not apply a voltage, this light is blocked and is denoted as the state "0".
Thus, through the integration and combination of hundreds of millions of optoelectronic chip components, the chip can store hundreds of millions of "0" or "1" data, by processing and calculating these data, can achieve a variety of functions.
However, the optoelectronic chip described here is only a simplified model of one of the many more complex and elaborate structures that can be found in real chips.
After reading this, some of you may ask: we have developed a good set of traditional chip technology, why do we need to study optoelectronic chips?
In fact, the answer to this question is very complex, but it is clear that as the traditional chip is becoming smaller and smaller, the integration degree (number of transistors per unit area) is becoming higher and higher, the processing process is becoming more and more complex, its performance has almost been promoted to the limit, it is difficult to have higher technology beyond in the future.
However, the demands on chip performance are almost endless. People in the future are bound to be more dependent on electronic devices than we are today, and traditional chips will one day be unable to meet People's Daily needs. So, with photoelectrons