chip Related Articles

This article would introduce MCU in details, including analysis its internal structure, and elaborate some important concepts, especially would put emphasis on the concept of memory decoding.   Catalog I. What is MCU? II. Some Basic Concepts 2.1 The Meaning of Rom 2.2 The Meaning of Bit 2.3 The Meaning of Bytes III. The Working Principle of Memory IV. MCU Circuit v. Memory Decoding FAQ   I. What is MCU?   MCU(microcomputer) is an integrated circuit chip. It integrates the microprocessor(CPU), which has data-handling technology such as arithmetic, logic and data transfer, etc, random access data memory(RAM), read-only program memory(ROM), input and output circuit (I/O port) that using the very large scale processing-data technology and may also include a timing counter, serial communication port (SCI), display drive circuit (LCD or LED drive circuit), pulse width modulation circuit (PWM), analog multiplexer and A/D converter, which form a minimum but perfect computer system.   Under the control of software, these circuits can complete the tasks specified by the program designer accurately, quickly, and efficiently. From this point of view, the single-chip microcomputer has the function which the microprocessor does not have, it has intelligent control functions which the modern industry control request separately. And this is the single-chip microcomputer's biggest characteristic.       II. Some Basic Concepts   2.1 The Meaning of Rom Let's think about a problem: when we write instruction in a programmer into an MCU and then take off it, the MCU can execute the instruction, so the instruction must be stored somewhere in the MCU. And this place can still maintain this instruction not to be lost after it power-off. What place is this? This place is the internal ROM of MCU, which is the read-only program memory. Why do you call it read-only memory? We use the programmer, external equipment, to write to the ROM operation under special conditions. In the MCU normal working conditions,  the data can only read but can’t write in, so we call it ROM.   2.2 The Meaning of Bit From the experiment above, we already know that the level of a lamp or a line can represent two states: 0 and 1. In fact, this is a binary bit, thus we call a line a bit, expressed in BIT.   2.3 The Meaning of Bytes A line can represent 0 and 1, two lines can express 00, 01, 10, 11 four states, that is, it can express 0 to 3, and three can express 0 to 7. The computer usually put with eight lines together, counting at the same time, can represent 0 to 255, for a count of 256 states. These eight lines or 8-bit is called a byte (BYTE).   III. The Working Principle of Memory   Structure All the instructions that a single-chip microcomputer can execute are the instructional systems of it. Different kinds of single-chip computers have different instructional systems. In order for a single-chip microcomputer to automatically complete a specific task, the problems to be solved must be programmed into a series of instructions (these instructions must be recognized and executed by the selected single-chip microcomputer). These instructions integrated into the program, and the program needs to be stored in memory—a storage unit.    The memory consists of many storage units (the smallest unit of storage), just as a building has many rooms, each room in a large building is assigned a unique room number. Each storage unit must also be assigned a unique address number, which is known as the address of the storage unit so that the address of the storage cell is known. The instructions are stored in these units. The storage unit can be found, where the stored instructions can be taken out and then executed.   Memory is the place where data is stored. It uses the electricity level to store the data, that is, it actually stores the electrical level, not the number of 1234 that we are used to thinking of. A memory is like a small drawer. If there are eight small drawers in a small drawer, each one is used to store the "charge," and the charge is passed in or released through the wire attached to it. You can think of a wire as a pipe, and the charge in the grid is like water, so it's easy to understand it. Each small drawer in memory is a place for data, which we call a ''bit''.   With this structure, we can start storing data. If we want to put in a data 12, that is 00001100, and we just have to fill the second and third squares with the charge, and the other cells are free of the charges. But the problem is that memory has a lot of cells, and the lines are parallel, and when you put the charge in it, you put the charge in all the cells, and when you release the charge, you release the charge from each cell. In the case of it, no matter how many cells the memory has, it can only be put in the same number, which is certainly not what we want.    A little bit to change structurally,  there's a control line on each unit, and if you want to put the data in the unit, give a signal to the control line of the unit. Therefore, the control line turns on the switch so that the charge can flow freely. And there is no signal on the other unit control lines, so the switch turns off and will not be affected, so that if you handle the control lines of different units, you can write different data to each unit. Similarly, if you want to take data from one unit, just turn on the corresponding control switch.     IV. MCU Circuit   A circuit is always made up of components connected by wires. In analog circuits, wiring is not a problem, because there is usually a serial relationship between the devices, and there are not many connections between the devices, but the computer circuits are different. The microprocessor is the core for it, each device must be connected to the microprocessor, the work of each device must be coordinated, so it needs a lot of connections.   If still like analog circuits, there will be an amazing number of lines between microprocessors and devices, so the concept of a bus has been introduced into the microprocessor, and each device has shared the connection. All 8 data lines are connected to eight common lines, that is, the equivalent of each device is in parallel, but this is not enough. If there are two devices delivering data at the same time, one is 0 and the other is 1, what exactly does the receiver get? This situation is not allowed, so control through the control line to make the device working time-sharing, at any time there can be only one device to send data ( multiple devices can receive at the same time).      V. Memory Decoding   So how do we control the control lines of each unit? It is not that simple to lead the control lines of each unit out of the integrated circuit. There are 65,536 units in a model 27512 memory, and if each line is drawn out, the integrated circuit must have more than 60,000 feet, so it is necessary to find a way to reduce the number of lines. We have a way called decoding, briefly introduce: one line can represent two states and two lines can represent four states and three lines can represent eight kinds, and so on, thus we only need 16 lines to represent 65536 states.   Since the decoding problem solved, let's focus on another problem. Where did the eight lines in each unit come from? Actually, it is connected to the computer, in general, the eight wires not only for memory but also connected to other devices. The problem arises in this way. Because these eight wires are not dedicated to the memory and the computer, it is not good if a unit is always connected to the eight wires. For example, if the value in this memory cell is 0FFH but there one unit is OOH, then what the line set at a high level or a low level?   Thus we have to separate them. The solution is: when the outside wire is connected to the pin of the integrated circuit, it does not directly attach to the units, but a set of switches is added to the middle. Normally we leave the switch off, and if we really want to write data to this memory, or read the data out of the memory, just turn the switch on. This set of switches is selected by three leads: read control, write control, and chip selector.    To write data into the chip, select the chip first, then send a write signal, the switch turns on, and the incoming data (charge) is written into the film chip. If you want to read, select the film chip first, and then send out the read signal, the switch turns on, and the data is sent out. The read and write signals are also connected to another memory at the same time, but the chip selector ends are different.   Although there is a read or write signal, there is no chip selection signal, so the other memory will not "misunderstand" and result in a conflict. What will happen if you pick two chips at the same time? Actually, this can’t be happening because the system is designed and controlled by computer, not by the human. If any, there’s something wrong with the circuit.     From the introduction above, we have seen that the eight lines used to transmit data are not dedicated, but shared by many devices, so we call it data bus. The data line of the device is called the data bus, and all the control lines of the device are called the control bus. There are memory cells in the internal or external memory and other devices of a single chip. Units must be assigned addresses before they can be used. Of course, the assigned addresses are also given in the form of electrical signals. Because there are too many memory cells, there are many lines for address allocation, which are called address buses. Sixteen address lines are also connected, called address buses.   FAQ   1. What are the characteristics of microcomputer? a. Small size and low cost. b. One user. c. Easy to use. d. Low computing power. e. Commonly used for personal application.   2. What are the advantages of microcomputer? a. This computer is widely used today. b. The microcomputer is small in size. c. The microcomputer is used to design different software and app. d. This type of computer is a low cost, so all the users can easily buy. e. No need for highly trained staff for operating microcomputer to office work.   3. Why microcontrollers are often called single chip computers? Single-chip computers are mainly of the form known as Microcontroller chips (the most commonly known are the PIC range by Microchip inc) and used in embedded devices. They provide much more basic functionality but are far simpler to work with as they don't require any external chips in order to function.   4. What is single chip microcomputer that has everything inbuilt? This is a microcomputer built using separate components (CPU, Memory, etc.). ... For some specific applications, we also have single chip computers in a VLSI chip. This single chip microcomputer will have a CPU, memory and I/O interfaces, timers, ADC/DACs etc. on a single chip itself.   5. What is difference between microprocessor and microcomputer? The main difference between Microprocessor and Microcomputer is that the Microprocessor is a computer processor contained on an integrated-circuit chip and Microcomputer is a small, relatively inexpensive computer. ... Microprocessors contain both combinational logic and sequential digital logic.   6. Is Raspberry Pi a microcomputer? The Raspberry PI is a microcomputer that's often used by hobbyists to create projects like animated LED displays or bird watchers.   7. Which is a feature of a single chip microcomputer? A single-chip microcomputer is a major branch of a microcomputer. The biggest feature of the structure is that the CPU, memory, timer and various input/output interface circuits are integrated on a very large-scale integrated circuit chip. In terms of its composition and function, a single chip is a computer.   8. What are the components of microcomputer? The main components are: (1) the central processing unit (CPU), (2) input devices, (3) output devices, and (4) memory. The CPU of a microcomputer performs all the arithmetic, logic, and data handling functions of the microcomputer.   9. Is microcontroller a microcomputer? A Microcontroller is a small and low-cost microcomputer, which is designed to perform the specific tasks of embedded systems like displaying microwave information, receiving remote signals etc.   10. What is the definition of microcomputer? Microcomputer, an electronic device with a microprocessor as its central processing unit (CPU). Microcomputer was formerly a commonly used term for personal computers, particularly any of a class of small digital computers whose CPU is contained on a single integrated semiconductor chip.   You May Also Like Transformers Basics: Construction, Types, Materials and Design Switched Mode Power Supply Tutorial: Principles & Functions of SMPS Circuits List of Basic Electronic Components Switching Power Supply Tutorial: 4V～16V

Catalog Introduction Design Flow of Chip Design Specification Development Design Details of the Chip Draw a Blueprint for the Plane About Wafer What Is a Wafer How to Make Single Crystal Wafer Metallurgical Purification Pulling the Crystal Design Flow of Chip Manufacture What Is an IC Chip Metal Sputtering Coating Photoresistance Etching Technology Photoresist Removal Nano-Process What Is the Nano-Process How Tiny Is the Nanometer Purpose of Reducing the Process Physical Limitations of Downsizing About Encapsulation Two Common Packages DIP Package BGA Package Two Ways to Reduce Size SoC SiP Introduction A chip is a silicon chip that contains an integrated circuit, so the chip is also called an integrated circuit. It may be only 2.5 centimeters in square size, but it contains tens of millions of transistors. Simpler processors, on the other hand, may have thousands of transistors engraved on chips which are a few millimeters in size. Chip is the most important part of electronic equipment, which undertakes the function of operation and storage. Design Flow of Chip The birth of a chip can be divided into two parts: design and manufacture. First, let's take a look at the complex and tedious chip design process. Fig 1. The process of making a chip is like building a house with Lego. First, the wafer is used as the foundation and the necessary IC chips can be produced after layers are stacked on top of each other. However, there is no use in having no amount of manufacturing capacity without a design drawing. Therefore, the role of an architect is very important. But who is the architect in IC's design? The next step is to introduce the IC design. In the IC production process, IC is mostly planned and designed by professional IC design companies, such as MediaTek, Qualcomm, Intel and other well-known large factories, all of which design their own IC chips to provide different specifications and efficiency chips for downstream manufacturers to choose from. Because IC is designed by the factories themselves, so IC design depends very much on the technology of engineers and the quality of engineers affects the value of an enterprise. But what are the steps engineers take to design an IC chip? The design process can be simply divided into the following steps. Design Specification Development In IC design, the most important step is specification development. This step is like deciding how many rooms, bathrooms, what building codes to comply with, and designing after all the features have been identified so that no additional time is spent on subsequent modifications. The IC design needs to go through similar steps to ensure that the chip is designed without any errors. The first step in specification development is to determine the purpose and effectiveness of IC and to set the general direction. The next step is to see what protocols to comply with, such as the wireless card chip needs to comply with IEEE 802.11 and other specifications. Otherwise, the chip will not be compatible with the products on the market, so that it will not be able to connect to other devices. Finally, the implementation method of this IC is established, different functions are allocated into different units, and the method of connecting different units is established, so that the specification can be completed. Design Details of the Chip After designing the specifications, it is followed by the details of the design chip. This step is like making a preliminary note of the planning of the building and depicting the overall outline for subsequent drawing. In IC chip, the hardware description language (HDL) is used to describe the circuit. The commonly used HDLs are Verilog, VHDL, and so on, which can easily express the function of a IC by code. This is followed by checking the correctness of the program's functionality and continuously modifying it until it meets the desired functionality. Fig 2. Verilog Example of 32 Bits Adder Draw a Blueprint for the Plane With a complete plan, the next step is to draw a blueprint for the plane. In IC design, the step of logic synthesis is to put the unmistakable HDL code into the electronic design automation tool (EDA tool), to let the computer convert HDL code into logic circuit, resulting in the following circuit diagram. After that, it is repeatedly determined whether the logic gate design conforms to the specification and is modified until the function is correct. Fig 3. The Result of the Synthesis of the Control Unit Finally, the synthesized code is put into another set of EDA tool for circuit layout and winding (Place And Route). After continuous detection, the following circuit diagram will be formed. You can see blue, red, green, yellow and other different colors, each of which represents a mask. As for the use of the mask, how should it be used? Fig 4. The Commonly Used Calculus Chip-FFT Chip, Which Completes the Circuit Layout and the Winding Result ——The chip is stacked by layers of masks. First of all, it is now known that an IC will produce multiple masks. These masks have the difference between the upper and lower layers and each layer has its own task. The following figure is a simple mask example. Taking the most basic element CMOS in the integrated circuit as an example, the full name of CMOS is complementary metal oxide semiconductor. That is, the combination of NMOS and PMOS to form CMOS. As for what is a metal oxide semiconductor (MOS)? This kind of component which is widely used in the chip is more difficult to explain, and it is more difficult for the general reader to figure it out, so there is no more detailed study here. In the following figure, on the left is the circuit diagram formed after the circuit layout and winding, and you have already known that each color represents a mask. On the right is the way each mask is spread out. Production is to start from the bottom, in accordance with the method proposed in the manufacture of the IC chip, layer by layer, and finally the desired chip will be produced. Fig 5.  At this point, you should have a preliminary understanding of the IC design. The overall view is very clear that IC design is a very complex major, but also thanks to the maturity of computer-aided software, so that IC design can be accelerated. The IC design relies heavily on the wisdom of engineers, and each of the steps described here has its own expertise and can be separated into multiple professional courses. For example, writing a hardware description language does not simply require familiarity with the programming language. You also need to understand how logic circuits work, how to convert the required algorithms into programs, and how synthetic software converts programs into logic gates. What Is a Wafer? In semiconductor news, it is always mentioned in the size of the wafer, such as 8-inch or 12-inch wafer. But what is the so-called wafer? What part of it is 8 inches? What is the difficulty of producing large wafers? Here is a step-by-step introduction to the most important foundation of semiconductors-what is a "wafer". Wafer is the basis for making all kinds of computer chips. We can compare chip manufacturing to building a house with Lego blocks and building the shape we want (that is, all kinds of chips) by stacking one layer after another. However, if there is no good foundation, the built house will be tilted back and forth, contrary to our wishes. In order to make the perfect house, we need a smooth substrate. For chip manufacturing, this substrate is the wafer that will be described next. First of all, think back to when you were a child playing with Lego blocks, there would be a small round bulge on the surface of the building blocks. With this structure, we can stack the two blocks firmly together without using glue. Chip manufacturing, also in a way like this, binds subsequent atoms to the substrate. Therefore, we need to find a substrate with a neat surface in order to meet the conditions needed for subsequent manufacturing. Fig 6. In solid materials, there is a special crystal structure. That is, single crystal (Monocrystalline). It has the characteristics of atoms one after another closely arranged together, which can form a flat atomic surface. Therefore, using single crystal to make wafer can meet the above needs. However, how to produce such a material? There are two main steps, respectively, purification and crystal pulling. After this, such a material can be completed. How to Make Single Crystal Wafer? Metallurgical Purification The purification is divided into two stages. The first step is metallurgical purification. During this process, we add carbon and convert silicon oxide into silicon with a purity of more than 98% in a redox manner. Most metals, such as iron or copper, are refined in this way to obtain sufficient purity of metal. However, 98% is still not enough for chip manufacturing and still needs to be further improved. Therefore, Siemens process will be used for purification, so that the high purity polysilicon needed for semiconductor process will be obtained. Fig 7. Silicon Column Manufacturing Process Pulling the Crystal Then there is the step of pulling the crystal. First, the high purity polysilicon obtained earlier is melted to form liquid silicon. After that, the single crystal silicon seed is in contact with the liquid surface and slowly pulls up as it rotates. As for why single crystal silicon is needed, that is because silicon atoms are arranged in the same way as people queue up. They will need to arrange the head so that later people can arrange it correctly. And silicon seed is an important row head, so that the later atoms know how to queue up. Finally, after the silicon atoms leaving the liquid surface solidify, the neatly arranged single crystal silicon columns are completed. Fig 8. Single Crystal Silicon Column But what do 8 inches and 12 inches stand for? It refers to the diameter of thin wafers being treated and sliced into，which is from the surface of the part of a crystal column that looks like a pencil rod. What is the difficulty of making large wafers? As mentioned earlier, the crystal column is made as if it were making marshmallows, rotating and forming at the same time. If you have made marshmallows, you should know that it is very difficult to make large and solid marshmallows, and the same is true of the crystal pulling process. The speed of rotation and the control of temperature will affect the quality of the crystal column. As a result, the larger the size, the higher the speed and temperature requirements are, so it is more difficult to make high-quality 12-inch wafers than 8-inch wafers. However, a whole silicon column cannot be made into a chip-making substrate. In order to produce a silicon wafer, the silicon column needs to be cut transversely into a wafer with a diamond knife, and the wafer can be polished to form the silicon wafer needed for chip manufacturing. After so many steps, the fabrication of the chip substrate is complete, and the next step is to stack the house, that is, chip manufacturing. So, how to make a chip? Manufacture ——Stacked chips After introducing what silicon wafers are, you also know that making IC chips is like building a house with Lego blocks, creating the shape you want by stacking layer after layer. However, there are quite a few steps to build a house, and so is IC manufacturing. What are the steps to make IC? Next, the process of IC chip manufacturing will be introduced. What Is an IC Chip? Before we begin, we need to know what an IC chip is. IC, which means integrated circuit (Integrated Circuit), is the design of the circuit that is in the form of stacking together. In this way, we can reduce the area required to connect the circuit. The following figure is a 3D diagram of the IC circuit, from which you can see that its structure is like the beams and columns of a house. It is done layer by layer and this is the reason why IC manufacturing is compared to building a house. Fig 9. 3D Profile of IC Chip From the 3D profile of the IC chip in the image above, the dark blue part at the bottom is the wafer introduced in the previous step. From this picture, we can see more clearly how important the wafer substrate plays in the chip. As for the red and khaki parts, they are the places to be completed when IC is made. First of all, the red part can be compared to the hall on the first floor of the building. The hall on the first floor is the door of a house because everyone and everything come in and out of here. It has more functionality under the control of traffic. Therefore, compared with other floors, the construction will be more complex and requires more steps. In IC circuit, this hall is the logic gate layer; it is the most important part of the whole IC by combining a variety of logic gates together and completes the fully functional IC chip. The yellow part is like a normal floor. Compared with the first floor, there will not be much complex structure, and each floor will not change much when it is built. The purpose of this layer is to connect the logic gates of the red part. The reason why so many layers are needed is that there are so many lines to be connected that a single layer cannot hold all the lines. So it is necessary to stack a few more layers to achieve this goal. Among them, the lines of different layers will be connected up and down to meet the needs of the wiring. ——Layered construction, layer by layer architecture Once you know the construction of IC, let's show you how to make it. Imagine that if we want to make a fine drawing with a paint spray tank, we need to cut out the cover plate of the figure and cover it on paper. Then spray the paint evenly on the paper and remove the mask when the paint is dry. After repeating this step over and over again, you can complete neat and complex graphics. IC is made in a similar way, by covering up a layer of stacking. Fig 10.  When making IC, you can simply divide into the above four steps. Although the actual manufacturing steps will be different and the materials used will be different, but generally using a similar principle. This process is slightly different from painting: IC manufacturing is to paint first and then cover while painting is to cover and then paint. And the processes are described below. Metal sputtering:  Sprinkle the metal material which is to be used evenly on the wafer to form a thin film. Coating photoresistance:  First put the photoresist material on the wafer, and then hit the beam on the desired part through the mask to destroy the structure of the photoresist material. Next, use chemicals to wash away the damaged material. Etching technology:  The silicon wafer without photoresistance protection will be etched by ion beam. Photoresist removal:  Use the photoresist solution to dissolve the remaining photoresist, so that a process can be completed. Finally, a lot of IC chips will be completed on a whole wafer, and then as long as the completed square IC chips are cut off, they can be sent to the packaging factory for packaging. What is the packaging factory? We'll have to explain it later. Nano-Process What is the nano-process? Samsung and TSMC compete fiercely in advanced semiconductor processes because both of them want to take the lead in wafer contract manufacturing to win orders, which has almost become a battle between 14 nanometers and 16 nanometers. But what is the meaning of 14 nm and 16 nm, and where do they refer? What are the benefits and problems that will be brought about by the reduction of the process? Next we will give a brief description of the nano-process. How tiny is the nanometer? Before you start, you need to understand what nanometer really means. Mathematically, nanometers are 0.000000001 meters, but this is a pretty bad example. After all, we can only see a lot of zeros after the decimal point, but we don't actually feel it. If you compare it with the thickness of nail, it may be more obvious. If you actually measure it with a ruler, you can tell that the thickness of the nail is about 0.0001 meters (0.1mm), that is to say, try to cut the side of a nail into 100000 lines, each of which is about one nanometer. From this, we can slightly imagine how tiny a nanometer is. Purpose of Reducing the Process After knowing how small the nanometer is, it is necessary to understand the purpose of reducing the process. The main purpose of reducing the transistor is to insert more transistors into smaller chips so that the chip will not become larger as a result of technological advances; second, it can increase the computational efficiency of the processor; moreover, reducing the volume can also reduce the power consumption. Finally, after the chip size is reduced, it is easier to plug into the mobile device to meet the needs of thinness and lightness in the future. Come back to explore what the nano-process is and we will take 14 nm as an example. The process refers to the minimum size of 14 nm in the chip. The following figure shows the appearance of a traditional transistor, as an example. The main purpose of reducing transistor is to reduce power consumption, but which part needs to be reduced to achieve this goal? The L in the figure on the left is what we expect to shrink. By reducing the gate length, the current can be routed from the Drain side to the Source end in a shorter path (if you are interested, you can use Google to search for MOSFET, which will be explained in more detail). Fig 11. In addition, computers operate on 0 and 1. How can we use transistors to meet this purpose? The way to do this is to determine whether the transistor has current flow. When a voltage supply is made at the Gate (green square), the current will flow from the Drain to the Source, and if there is no supply voltage, the current will not flow, so that it can represent 1 and 0. (As to why 0 and 1 are used to judge, if you are interested, you can go to the Brin algebra. That is the way we use this method to make a computer.) Physical Limitations of Downsizing However, the process cannot be reduced indefinitely. When we narrow the transistor to about 20 nanometers, we will encounter problems in quantum physics, so that the transistor has a leakage phenomenon, offsetting the benefits of L. As a way to improve, the concept of FinFET (Tri-Gate) was imported, as shown in the figure above. The leakage caused by physical phenomena can be reduced by importing this technology. Fig 12. More importantly, this method can increase the contact area between the Gate end and the lower layer. In traditional practice (top left), the contact surface has only one plane, but with FinFET (Tri-Gate), the contact surface will become three-dimensional, and the contact area can be easily increased. This allows the Source-Drain side to be smaller while maintaining the same contact area, which is of considerable help in reducing the size. Finally, why would anyone say that it would be a pretty serious challenge for factories to enter the 10-nanometer process? It is mainly because the size of an atom is about 0.1 nanometers, and in the case of 10 nanometers, there are fewer than 100 atoms in a line. It is very difficult to make, and as long as there is an atomic defect, such as atoms falling out or impurities in the production process, there will be unknown phenomena, affecting the yield of the product. If you can't imagine the difficulty, you can do a small experiment. Line up a 10 × 10 square with 100 small beads on the table, cut a piece of paper to cover the beads, then brush off the beads next to it with a small brush, and finally make it form a 10 × 5 rectangle. In this way, we can know the difficulties faced by the major factories and how difficult it is to achieve this goal. Encapsulation After a long process, from design to manufacture, finally we got an IC chip. However, a chip is so small and thin that it can be easily scratched and damaged if it is not protected from the outside. In addition, because of the small size of the chip, if you do not use a larger size of the shell, it will not be easy to manually place on the circuit board. Therefore, the next step is to describe the encapsulation: Two Common Packages At present, there are two common packages; one is the DIP package, which is common in electric toys and looks like a centipede, the other is the BGA package, which is common when buying boxed CPU. As for other packaging methods, there are PGA (Pin Grid Array) used in the early CPU or an improved version of QFP (plastic square flat package) of DIP. Because there are so many packaging methods, only DIP and BGA encapsulation are described below: ——Enduring Traditional Packaging DIP Package The first thing to introduce is the Dual Inline Package (DIP), we can see from the following figure that the IC chip with this package will look like a black centipede at the foot of the dual inline connection and this is the earliest IC packaging technology. It has the advantage of low cost and is suitable for small chips without too many wires. However, because most of them are plastic, the heat dissipation effect is poor, which cannot meet the requirements of the current high-speed chips. Therefore, most of the chips using this package are durable chips, such as OP741 shown in the following figure or smaller IC chips with less speed requirements and fewer holes. Fig 13.  The IC chip shown on the left is a common voltage amplifier named OP741. On the right is its section. The package connects the chip to the leadframe with a gold wire. BGA Package As for spherical array (Ball Grid Array,BGA) packaging, compared with DIP, it is smaller and can be easily placed in smaller devices. In addition, because the pin is located under the chip, it can hold more metal pins than the DIP so it is Ideal for chips that require more contacts. However, the cost of this packaging method is high and the connection method is more complex, so it is mostly used in high unit price products. Fig 14.  On the left is a chip encapsulated in BGA. On the right is a schematic diagram of BGA using a cladding packaging. ——The rise of mobile devices and the emergence of new technologies on the stage Two Ways to Reduce Size However, the use of these packaging methods will cost a considerable amount of volume. For example, today's mobile devices, wearing devices, and so on, require quite a variety of components. If each component is packaged independently, it will cost a lot of space. Therefore, there are two ways to meet the requirements of reducing size. They are SoC (System On Chip) and SiP (System In Packet). SoC At the beginning of the rise of smart phones, the term SoC can be found in major financial magazines, but what is SoC? To put it simply, ICs with different functions are integrated into one chip. By this method, not only the volume can be reduced, but also the distance between different IC can be reduced, and the calculation speed of the chip can be improved. As for the manufacturing method, during the IC design phase, different ICs are put together and then a mask is made through the design process described earlier. However, SoC is not the only advantage; to design a SoC requires considerable technical cooperation. When IC chips are encapsulated, they have their own external protection, and the distance between IC and IC is long, so there is no interactive interference. But when all the ICs are wrapped together, it is the beginning of a nightmare. The IC design factory has to change from the original simple design IC, to the IC which requires them to understand and integrate the various functions. Therefore, it increase the workload of engineers. In addition, there will also be a lot of situations, such as the high-frequency signal of the communication chip may affect the IC of other functions and so on. In addition, SoC also needs to obtain IP (intellectual property) authorization from other vendors in order to put components designed by others into SoC. Because making SoC needs to obtain the design details of the whole IC in order to make a complete mask, which also increases the design cost of SoC. Some people may question why not just design one by yourself. That is because designing all kinds of IC requires a lot of knowledge related to the IC, only a rich enterprise like Apple can have a budget to poach top engineers from well-known enterprises. It's still a lot cheaper to design a whole new IC through collaborative licensing than to develop it by yourself. SiP As an alternative, SiP has leapt onto the stage of integrating chips. Unlike SoC, it buys IC from different enterprises and finishes the last step, which is to encapsulate the IC. In this way, the IP licensing step is eliminated and the design cost is significantly reduced. In addition, because they are independent ICs, the degree of interference with each other is greatly reduced. Fig 15. Apple Watch uses SiP technology to package the entire computer architecture into a chip, not only to meet the desired performance but also to reduce the size, so that the watch has more space for battery release. The most famous product using SiP technology is Apple Watch. Because the internal space of Watch is too small, it cannot use the traditional technology, the design cost of SoC is too high, SiP has become the first choice. With SiP technology, not only the volume can be reduced, but also the distance between each IC can be shortened, so SiP can be a feasible compromise. The following figure shows the structure of the Apple Watch chip, and you can see that quite a few IC are included in it. Fig 16. Internal configuration Diagram of S1 Chip encapsulated by SiP in Apple Watch After the packaging is completed, we will enter the testing stage. At this stage, it is necessary to confirm whether the encapsulated IC is functioning properly and that it can be shipped to the assembly plant after it is correct, so that the electronic products we can see can be made. So far, the semiconductor industry has completed the task of the whole production.

SummaryDo you think that electornics and light one day can go well toghther on a standard‘CMOS' chip one day? It's reported that researchers have succeeded in introducting a light connection into the heart of a semiconductor chip a few days. In this way,two circuits can communicate.Or: the worlds of electronics and photonics are connected. BodyWhat is particularly attractive is researcher of the university of Twente--Satadal Dutta's solution--A light connection into the heart of a semiconductor chip. There is no special materials or manufacturing processes are needed: the light comes from silicon. The light source, detector and the light channel can be made using the technology that is used to make the electronic circuits. Fully optical circuits are available nowadays, but they use materials like indium phosphide and gallium arsenide, which can't easily be combined with the CMOS chip processes used for semiconductor chips you'll find in today's smartphones.  Connecting WorldsThere is a predictions say that all-optical circuits may become the‘new electronics'. In the transition from electronic to optic circuits, hybrid circuits, like the one Dutta designed, could play an important role. Satadal Dutta (1990, Barrackpore, India) did his PhD research in the Semiconductor Components group of Prof Jurriaan Schmitz, together with the Integrated Circuit Design group of Prof. Bram Nauta. Dutta defended his thesis 'Avalanche-mode silicon LEDs for monolithic optical coupling in CMOS technology' on 8 November. It was supported financially by NWO-TTW in The Netherlands and by NXP Semiconductors.  Avalanche LEDThe alternative would be: make a LED out of silicon. And that's the problem: silicon only emits a tiny amount of infrared light, while a detector made out of silicon needs visible light. They are talking and listening at different wavelengths. Dutta therefore chooses a remarkable way out: connect the LED reverse. At low voltages, there's no current, but at a voltage that is high enough, there will be a small current that amplifies itself like an avalanche. In this 'avalanche mode', the LED will transmit visible light. Using the same process, the light detector, as well as the light channel in-between can be made. Thanks to the special comb structure that Dutta designed, the light source gets more uniform and energy efficient. IsolationAn optical link on a chip is a good way to 'galvanically' isolate two circuits from each other. This is often necessary in cases where one circuit is a low-voltage and low-current one, while the other is a high-power circuit. They should be connected, but not by conducting wires, for reasons of safety. A classic transformer is an option then, but an optical connection is often used as well. Until now, this is a separate 'optocoupler', which is large and has a limited bit rate. Dutta's new solution is much more compact as an alternative: it total, it is just a few tens of microns and it offers the protection that's needed. Compared to optical channels in full-optical circuits, the energy consumption is relatively high, as there is quite some scattering of light. On the other hand: designing the electronics around the optical link in an efficient way, the amount of light needed for a successful connection, can be kept to a minimum. 

As the development of social technology,computer is becoming an universal things in families.The huge increase in computing performance in recent decades also has been achieved by squeezing ever more transistors into a tighter space on microchips. However,it's the tighter space on microchips that leads to effects such as signal leakage betwwen components,which will slow down communication between different parts of the chip. People in technology call this kind of delay as " interconnect bottleneck" as it is becoming an increasing problem in high-speed computing system.  Researchers are trying their best and all their professional knowledge to consider how to solve this problem. According to Pablo Jarillo-Herrero,an associte professor of physics at MIT,however,one way to tackle the interconnect bottleneck is to use light rather that wires to communicate between different parts of a microchip.But it is a hard work that the material "sillicon" used to build chips,does not emit light easily. The article about a light emitter and detector that can be integrated into sillicon CMOS chips was published in the Journal Nature Nanotechnology,a monthly peer-reviewed scientific journal published by Nature Publishing Group.This paper was written by MIT postdoc Ya-Qing Bie who joined Jarillo-Herrero and an interdisciplinary team including Dirk Englund, an associate professor of electrical engineering and computer science at MIT . According to this paper,the device is built from a semiconductor material called molybdenum ditelluride. This ultrathin semiconductor belongs to an emerging group of materials known as two-dimensional transition-metal dichalcogenides.  Unlike conventional semiconductors, the material can be stacked on top of silicon wafers, Jarillo-Herrero says."Researchers have been trying to find materials that are compatible with silicon, in order to bring optoelectronics and optical communication on-chip, but so far this has proven very difficult,for example,gallium arsenide is very good for optics, but it cannot be grown on silicon very easily because the two semiconductors are incompatible." on the contrary,the 2-D molybdenum ditelluride can be mechanically attached to any material.Another difficulty with integrating other semiconductors with silicon is that the materials typically emit light in the visible range, but light at these wavelengths is simply absorbed by silicon.Molybdenum ditelluride emits light in the infrared range, which is not absorbed by silicon, meaning it can be used for on-chip communication.To use the material as a light emitter, the researchers first had to convert it into a P-N junction diode, a device in which one side, the P side, is positively charged, while the other, N side, is negatively charged. In conventional semiconductors, this is typically done by introducing chemical impurities into the material. With the new class of 2-D materials, however, it can be done by simply applying a voltage across metallic gate electrodes placed side-by-side on top of the material.Jarillo explained continuelly:""That is a significant breakthrough, because it means we do not need to introduce chemical impurities into the material [to create the diode]. We can do it electrically." Once the diode is produced, the researchers run a current through the device, causing it to emit light."So by using diodes made of molybdenum ditelluride, we are able to fabricate light-emitting diodes (LEDs) compatible with silicon chips." The device can also be switched to operate as a photodetector, by reversing the polarity of the voltage applied to the device. This causes it to stop conducting electricity until a light shines on it, when the current restarts,so that the devices are able to both transmit and receive optical signals. This device is a proof of concept and there are a great deal of work need to be done before the technology can be developed into a commercial product, Jarillo-Herrero says. The researchers are now investigating other materials that could be used for on-chip optical communication.Most telecommunication systems, for example, operate using light with a wavelength of 1.3 or 1.5 micrometers, howevermolybdenum ditelluride emits light at 1.1 micrometers. This makes it suitable for use in the silicon chips found in computers, but unsuitable for telecommunications systems. "It would be highly desirable if we could develop a similar material, which could emit and detect light at 1.3 or 1.5 micrometers in wavelength, where telecommunication through optical fiber operates,"  Jarillo-Herrero added. In the end,researchers are another ultrathin material called black phosphorus, which can be tuned to emit light at different wavelengths by altering the number of layers used. They hope to develop devices with the necessary number of layers to allow them to emit light at the two wavelengths while remaining compatible with silicon. The article was ended by Jarillo hopes that communication on-chip by optical signals instead of electronic signals because they can do so more quickly whle comsuming less power. Well,let's cheer researchers on and hope them get success in the near future. 

（The new device is smaller than a thumbnail with a size of 0.1 x 4mm, and could be integrated into everyday electronic devices like smartphones.） Integrated circuits, so called chips, are used in everyday electronic equipment like mobile phones and computers. It is a set of electronic circuits on one small flat piece of semiconductor material, normally silicon. But this material has some limitations when it comes to processing data. To overcome these limitations and improve data processing, researchers are developing optical circuits made of chalcogenide glass. This special type of glass is used for ultrafast telecommunication networks, transferring information at the speed of light. Integrating these glass optical circuits into silicon chips could lead to a more advanced communications system, processing data a hundred times faster. Can these two materials be combined? The answer is yes! In a collaboration with physicists in the University of Sydney's Australian Institute for Nanoscale Science and Technology (AINST), the Australian National University (ANU) and RMIT University, the CUDOS research group around PhD candidate Blair Morrison and senior researcher Dr Alvaro Casas Bedoya created compact, mass manufacturable optical circuits with enhanced functionalities by combining nonlinear glasses with silicon-based material. "In the last few years the group at the University of Sydney has repeatedly demonstrated exciting functionalities, such as broadband microwave devices that enhance radar, using these novel chalcogenide glasses," Blair Morrison said from the University of Sydney CUDOS node. "Now we have shown it is possible to combine this material with the current industry standard platform for photonic integration, silicon," he said. "We integrated a novel nonlinear glass into an industrially scalable CMOS compatible platform. We maintained the key advantages of both the silicon and the glass, and made a functional and efficient ultra-compact optical circuit," said Dr Alvaro Casas Bedoya who is the lead photonics nanofabrication manager for CUDOS. "A wealth of new opportunities will be created, and this takes us one step closer to moving our research from the lab into industrial applications," said Blair Morrison. CUDOS Director and ARC Laureate Fellow Professor Benjamin Eggleton from the University of Sydney said this new approach will one day allow the industry to miniaturise the photonics functionalities from devices that are the size of a laptop to the size of a smartphone and even smaller, allowing for deployment in real world applications. "This is exciting, because this is a platform which is more compatible with existing semiconductor manufacturing and will allow us to integrate multiple functionalities on a single silicon chip, with active and passive components, such as detectors and modulators, required for advanced applications," said Professor Eggleton who supervised the project. The multi-university research team went through the whole manufacturing process: The fabrication of these devices uses silicon wafers from a semiconductor foundry in Belgium, a dedicated facility in ANU's Laser Physics Centre for the glass deposition, lithography in the RMIT University's School of Engineering and are then characterised and tested in the University of Sydney's AINST. To showcase the potential of the new approach, the CUDOS researchers further demonstrated a compact novel laser based on the light-sound interactions, the first time in an integrated optical circuit. "The breakthrough here is this realisation that we can actually interface, we can integrate that glass onto silicon and we can interface from silicon to the glass very efficiently -- we can harness the best of both worlds," Professor Eggleton said. Professor Susan Pond, the Director of AINST, emphasized that this project is one of AINST flagship activities that deals with harnessing interactions between photons and phonon at the nanoscale. This work links fundamental research in light matter interactions at the nanoscale with an end user perspective and strong coupling to industry. Ref.KY32-LMX6502SQKY32-LN2300KY32-LN3251MPW

Today, at the imec technology forum (ITF2017), imec demonstrated the world's first self-learning neuromorphic chip. The brain-inspired chip, based on OxRAM technology, has the capability of self-learning and has been demonstrated to have the ability to compose music.The human brain is a dream for computer scientists: it has a huge computing power while consuming only a few tens of Watts. Imec researchers are combining state-of-the-art hardware and software to design chips that feature these desirable characteristics of a self-learning system. Imec's ultimate goal is to design the process technology and building blocks to make artificial intelligence to be energy efficient so that that it can be integrated into sensors. Such intelligent sensors will drive the internet of things forward. This would not only allow machine learning to be present in all sensors but also allow on-field learning capability to further improve the learning.By co-optimizing the hardware and the software, the chip features machine learning and intelligence characteristics on a small area, while consuming only very little power. The chip is self-learning, meaning that is makes associations between what it has experienced and what it experiences. The more it experiences, the stronger the connections will be. The chip presented today has learned to compose new music and the rules for the composition are learnt on the fly.It is imec's ultimate goal to further advance both hardware and software to achieve very low-power, high-performance, low-cost and highly miniaturized neuromorphic chips that can be applied in many domains ranging for personal health, energy, traffic management etc. For example, neuromorphic chips integrated into sensors for health monitoring would enable to identify a particular heartrate change that could lead to heart abnormalities, and would learn to recognize slightly different ECG patterns that vary between individuals. Such neuromorphic chips would thus enable more customized and patient-centric monitoring."Because we have hardware, system design and software expertise under one roof, imec is ideally positioned to drive neuromorphic computing forward," says Praveen Raghavan, distinguished member of the technical Staff at imec. "Our chip has evolved from co-optimizing logic, memory, algorithms and system in a holistic way. This way, we succeeded in developing the building blocks for such a self-learning system." Ref:KY32-MAX1490AEPG+KY32-LN3251MPW