Is the processor just the CPU small? Why is it used?

CPU is the first processor, a very small chip.

Professional explanation: CPU is the heart of the computer system, and the rapid development process of computers, especially microcomputers, is essentially the process of CPU development from low-level to high-level, from simple to complex.

First, the concept of CPU.

CPU (Central Processing Unit), also known as ** processor, its main function is to perform operations and logical operations, and the internal structure can be roughly divided into several parts, such as control units, arithmetic logic units and storage units. According to the word length of the information it processes, it can be divided into: eight-bit microprocessor, sixteen-bit microprocessor, thirty-two-bit microprocessor and sixty-four-bit microprocessor and so on.

Second, the main performance indicators of the CPU.

Main frequency: The clock frequency at which the core of the CPU works internally, generally in megahertz (MHz). This is one of the parameters that we usually care about the most whether we are using or buying a computer, and what we usually refer to as etc.

For the same type of CPU, the higher the frequency, the faster the CPU speed and the higher the performance of the whole machine.

External frequency and multiplier: The external clock frequency of the CPU is the external clock frequency of the CPU. The external frequency is provided by the computer motherboard, and the relationship between the main frequency of the CPU and the external frequency is: the main frequency of the CPU, the external frequency, and the number of octapers.

Internal cache: It is made with extremely fast SRAM, which is used to temporarily store the latest part of the instructions and data during CPU operation, and the access speed is the same as the CPU frequency, and the capacity of the internal cache is generally in KB. When it is working at full speed, the larger its capacity, the easier it is for the most frequently used data and results to enter the CPU as soon as possible for computing, and the less data can be exchanged between the CPU and the slower access external cache and memory, and the faster the computing speed can be increased relative to the computer.

Address bus width: The address bus width determines the physical address space that the CPU can access, which is simply how much memory the CPU can use.

Multi** Extended Instruction Set (MMX) Technology: MMX is a new technology adopted by Intel to enhance the Pentium CPU for audio, video, graphics and communication applications. This technology adds a new 57 MMX instructions to the CPU, which is about 60 more capable of processing more than ordinary CPUs when running programs containing MMX instructions.

Even programs that don't use MMX instructions can get a performance boost of about 15%.

Microprocessors have changed our lives in ways that were previously unimaginable for granted. In the sixties, computers were so big that they filled an entire room, and only a few people could use them. The invention of the integrated circuit in the mid-sixties made it possible to miniaturize the circuit on a single silicon wafer, laying the foundation for the development of microprocessors.

In the foreseeable future, the processing power of CPUs will continue to maintain rapid growth, and miniaturization and integration will always be the development trend, and at the same time, different levels of products will be formed, including dedicated processors.

Yes, he is the ** processor, he processes all data, he is the human brain, understand no.

The 2nd floor is powerful, and the CPU is the human brain.

Hehe, the friend on the 2nd floor is too excited to estimate.

Yes, a chip. It acts like the human brain. That's why it's called a dissipator.

Originally, the CPU is the human brain. Scold. All issues have to be left to him to make the final decision and deal with.

The CPU is cut from the wafer, and the size of the wafer is limited.

Therefore, the smaller the surface area of the CPU, the more CPU will be manufactured. the lower the cost.

Monocrystalline silicon wafers are made of ordinary silicon sand drawing, and are made into monocrystals through a series of measures such as dissolving, purifying and distilling.

The main purpose of CPU overclocking is to increase the working frequency of the CPU, that is, the main frequency of the CPU. The main frequency of the CPU is the product of the external frequency and the frequency doubling. For example, the external frequency of a CPU is 100MHz, and the frequency multiplier is , and its main frequency can be calculated External frequency and frequency doubling 100MHz = 850MHz.

You can increase the CPU clock speed by changing the CPU frequency doubling or external frequency. But if you're using an Intel CPU, you can ignore the multiplier, because the Intel CPU uses a special manufacturing process to prevent the multiplier from being modified. AMD's CPU can modify the frequency multiplier to return to the service, but modifying the frequency doubling does not improve the CPU performance as well as the external frequency.

That is, I also think that if it is made to the size of a brick, it can also defend yourself, hehe.

Joke this is because the wafer of the chip is cut from a large piece of silicon wafer, and the size is large, and the number of cuts is small, and the manufacturer will lose money.

The CPU size design is small for the following reasons:

1.Wafer specifications, new specifications mean new production equipment and production lines, expensive, and wafers within a certain quality range, can't be too big.

2.Timing and latency: If the CPU is too large, the latency between internal computing units will increase, and the latency of the cache will also increase, and the performance improvement curve will steepen, until it is not as good as multiple CPUs.

3.Heat, calorific index increases, not to mention air cooling, water cooling and oil cooling can't be reduced.

4.Energy consumption, when it reaches a critical point, increasing the size is greater than the energy consumption of multiple CPUs.

5.In the process, the wafer has the best physique in the center, and it is difficult to control the quality if it is the most inferior and the largest in the surrounding area, and the yield rate will drop sharply.

The CPU is cut from the wafer, and the size of the wafer is limited.

Therefore, the smaller the surface area of the CPU, the more CPU will be manufactured. The lower the cost, the monocrystalline silicon wafer is extracted from ordinary silica sand, and the monocrystalline silicon rod is made into a series of measures such as dissolution, purification and distillation, and the monocrystalline silicon rod becomes a wafer after polishing and slicing. Wafers are the most commonly used semiconductor materials, and are divided into 4 inches, 5 inches, 6 inches, 8 inches and other specifications according to their diameter, and recently developed 12 inches and even larger specifications (14 inches, 15 inches, 16 inches、......20 inches or more, etc.).

The larger the wafer, the more ICs can be produced on the same wafer, which can reduce costs. However, higher material technology and production technology are required.

CPU is a high-tech product, and there are many reasons why it is so small, such as: CPU is the heart of the whole computer, and it must have strong transmission and computing capabilities, so the production materials used are very expensive, and the small volume can reduce the cost; At the same time, the small size can reduce the interference of resistance, make the current fully used, and achieve rapid information transmission. Today's machines are all about easy to carry, and making them small also caters to this.

The main reasons are:

The CPU is the heart of the whole computer, and it must have strong transmission and computing capabilities, so the production materials used are very expensive, and the small volume can reduce the cost;

The small size can reduce the interference of the resistance, make the current fully used, and achieve rapid information transmission. Today's machines are all about easy to carry, and making them small also caters to this.

The small size will correspond to much lower power consumption and power supply, which will greatly help the overall performance of the computer.

At first, discrete transistors developed analog circuit designs, but their complex wiring led to complex circuit design, high volume and power consumption, and high design costs. To overcome these shortcomings, semiconductor processes were developed.

TTL and ECL are both digital logic structure circuits composed of transistors, and CMOS is a digital logic structure circuit composed of field effect transistors, and transistors and field effect transistors represent two manufacturing processes. It was proved that CMOS became the first choice for digital design because it can reduce the line width and improve the degree of integration.

Transistor design scale of integrated circuits:

10μm 2300 transistors

134 000 transistors

3 100 000 transistors (early).

7 500 000 transistors (late).

90nm 125 000 000 transistors

Now it's all 45nm and the number of transistors is in the billion range.

Moore's Law says that every 18 months, the transistor density of a chip doubles and the computing performance doubles.

In the pursuit of large computing and high-speed computing, the required performance continues to improve. Power consumption and heat generation are very important factors in chip design, and too much heat production can lead to calculation errors and even burnout.

NRE (Non-Recurrent Engineering) Cost Design cost, sample cost.

The cost of recurrent is directly proportional to the yield and proportional to the chip area.

Cost per IC = Recurrent Cost + NRE Cost Total Finished Product.

Overall, the cost per transistor is reduced.

The design cost (one-time investment) is very, very expensive! Not always successful.

The smaller the chip area, the more inevitable it is.

CPU is not an organ in the human body, not the bigger the better, just like the horsepower of the car, the fuel consumption will soar, but you just use this high-horsepower car to do something relatively simple, killing chickens with a knife?

Take a look at the CPU's heatsink and you'll find out. The volume ratio is quite large.

If the CPU is bigger, the heatsink will be too big.

And big is not necessarily good, improving the process and process is the key.

Modern CPUs are made using silicon and other metals and chemicals. Silicon has the properties of semiconductors, suitable for the manufacture of various small transistors, is now one of the main materials for large-scale integrated circuits, this high-purity monocrystalline silicon is not cabbage. Thank you!!!

It's the same as the size of the heart, it's enough, and the home computer, why bother, it's not big.

Bigger is costly and consumes a lot of power.

Power issues, but also portability, etc.

You see, now that technology is constantly developing, nothing is getting smaller、、、

First of all, there are two main manufacturers that produce processors, AMD and Intel.

The problem you mentioned is caused by a difference in the naming conventions between the two vendors.

2000+ and 3000+ are nominal values, meaning the equivalent of a processor with a main frequency of 2000MHz (that is, its actual frequency cannot be reached. It is AMD's practice to adopt the nominal value.

"Pentium" is a series of Intel processors that are positioned in the mid-to-high-end market. There is also the "Celeron" series for the low-end market, and the "Xeon" series for the server market. The current overwhelming dual-core and quad-core processors belong to the latest "Core" series.

Refers to the actual frequency.

Intel's processor naming convention consists of "series name" + "product code", and there are many products in the same series, and these products are distinguished by product code. For example, "Pentium 4

631" represents the processor frequency of the Pentium 4 core.

Addendum: I try to say it simply, not for points, as long as it can help.

MHz is a unit smaller than g, and their relationship is the relationship between "centimeter" and "meter".

2000MHz is equal to 2G

If you buy a 2-hand computer, try not to buy a machine before 04. The second-hand machines from 04 to the present are enough to meet your requirements, too early to guarantee it.

You can buy about 800 yuan to 1,000 yuan for 2 hand computers, and when you go to buy it, you can say to the profiteer that you want to "Ben 4."

The processor or the "3000+ processor of Athlon 64" is the mainstream configuration of the year, which can be said to be a guarantee of performance. Anything lower than these two can be disregarded.

The memory should be at least "512 megabytes", and "1GB" is better.

The graphics card should be 9550 or 5700 of "double 128", (which means that the video memory is 128 megabytes, and the video memory bit width is 128 bits).

Hard disk at least"80gb"Target.

How to see good and bad, this can't be said, sometimes very new hardware profiteers may have also moved. The only way is to ask the profiteer to guarantee that the package will be returned within a week if there is a problem when it is bought. After you buy it, you can turn it on for a week (at least three days!).

It's necessary! Conduct a high-intensity test to test the stability of the machine. Normal machine 7*24 open for a week will not be a problem.