Chapter 5-Pentium Core (Intel-Intel)

From: http://www.google.com.hk/ggblog/googlechinablog/2007/10/intel_8535.html

 

The west coast of the United States between San Francisco and San Jose City (San Jose), around the San Francisco Bay has dozens of kilometers long, several kilometers wide canyon, usually called Silicon Valley. (The Green coil in the figure) Silicon Valley is called not because it produces silicon, but because it has many semiconductor companies that use silicon, including Intel, the world's largest semiconductor company. More than half of the world's computers use its central processor (CPU). Its impact on our daily lives is comparable to that of few companies. In the previous chapter, we introduced Moore's theorem and Andy Gates's theorem. Moore is the founder of Intel, while Andy. grove is the fourth company to join intel and turn intel into the CEO of the world's largest semiconductor company. Today, Intel has nearly 100,000 people, with an annual output of 36 billion US dollars and a market value of 140 billion US dollars. For more than 30 years, the key to Intel's success was to take advantage of the wave of the PC revolution, especially the strong partner of Microsoft. Second, over the past 30 years, it has been improving the performance of its processor for PC users around the world in strict accordance with the astonishing high speed predicted by its founder. in its own words, it gives each microcomputer a Pentium Core.

1. Heroes of the Times

Intel was founded in Silicon Valley in 1968 by Gordon E. Moore and Robert Noyce. Previously, Moore and noisy founded Fairchild Semiconductor with six other people in 1956. Compared with companies such as IBM, Dec, and HP, Intel can be regarded as only one baby for a long time. It has two meanings: first, it is a small company with a small number of people and small businesses. Second, before 1980s, almost all computer companies such as IBM and Dec have their own central processors, so these computer companies represent the highest level of processor design and manufacturing, while Intel produces low-performance microprocessors, it is used to supplement the low-end market that big computer companies cannot see. In terms of performance, Intel's 1980s processor was not as good as IBM's 1970s processor, but its processor was quite affordable. Even when Intel produced the famous 16-bit 1970s processor by the end of 8086, we still regard it as a younger brother. For a long time, Intel products were considered low performance and low price. Although it is cost-effective, it is not a cutting-edge product.

Although 8086 was the ancestor of all our ibm pc processors today, even intel did not predict its reusability at the time. At that time, intel did not have a clear market position for 8086, but wanted to make as many promotions as possible. IBM was just one of Intel's numerous clients, big and small. In 1981, in order to quickly develop a PC, IBM was too lazy to design its own processor. It was used directly for Intel's 8086. At this moment, Intel became famous. In 1982, Intel developed the second-generation PC Processor 8086 that is fully compatible with 80286 and used on ibm pc/. Since IBM cannot stop other compatible machines, with the creation of the world's first compatible machine for ibm pc in 1985, compatible machine manufacturers have sprung up all over the world. These compatible machines have different hardware paths, but to be compatible with IBM PCs, the processors must be from Intel. Is the ecological chain of the entire PC industry.

It can be seen that in this ecosystem, only Microsoft, as the operating system developer, and Intel, as the processor manufacturer, are in an irreplaceable position. Therefore, the rise of Intel has become a historical necessity. This is exactly the trend of heroes.

Of course, although the wave of the information revolution has pushed intel to the forefront, intel must be able to lead the technological revolution in computer processors. Intel's CEO Andy Grove proved intel was the king in the face of opportunities and challenges. Intel started in 1980s just ten years ago in Japan. At that time, the total market value of the Japanese stock market accounted for half of the world. The total value of real estate near Tokyo is equivalent to half of the total value of the U.S. House market. The three largest semiconductor companies in the world are all in Japan, where Japanese CHiPs once accounted for 60% of the total number (Note: Not 60% of the price ). As a result, some Japanese politicians are blindly arrogant and think that when Japan fully challenges the United States, the world is wondering if the United States will lag behind Japan in semiconductor technology. But with a calm analysis, the semiconductor market around the world will find that Japan's semiconductor industry is concentrated on low-tech chips, such as memory chips (memory ), the world's high-end chip industry, such as computer processors and communication digital signal processors are all in the United States. In 1980s, Intel decisively stopped its memory business and gave the market to the Japanese, so it began to focus on processing. At that time, Japanese semiconductor companies made a lot of money throughout the city. Japan cheered that they beat Americans. In fact, this is just a move for US companies such as Intel to abandon their ambitions. In 1985, after Motorola, Intel developed a 32-bit microprocessor of 80386, which began to expand its market share in the entire semiconductor industry. The cost for developing this chip is more than $0.3 billion. Although it is far lower than the current development cost of Intel's new processor chip, it was indeed a great bet at the time, this research cost is several times more than China's total investment in semiconductor research in a five-year plan. Intel has completed the success of the ibm pc compatible machine market by 80386.

Next in 1989, Intel launched the transition product 80386 from 80486 to the Pentium processor, which is actually 80386 plus a floating point processor 80387 and cache ). With 80486 of sales, Intel surpassed all Japanese semiconductor companies and took the lead in the semiconductor industry. By the way, today, Japan's stock market is less than 1990 in 40%, but the United States has increased by five times. In 1993, Intel launched the Pentium processor. Since the beginning of the Pentium, Intel has no longer named its products in numbers, but in the industry and academia, we are still habitually referred to as Intel's processor x86 series.

The birth of the Pentium caused intel to get rid of the hat that will only be used as a low-performance processor. Since the speed of the Pentium processor has reached the level of the workstation processor, the high-end microcomputer began to replace the low-performance graphics workstation. Today, even Sun, the earliest workstation manufacturer, IBM, the world's largest computer company, and Apple, which never used Intel processors before, all started to use intel or Intel-compatible processors in their own computers. Now, Intel has monopolized the computer processor market.

2. Battle of Intel Motorola

After all, capital-intensive Japanese semiconductor companies cannot be competitors of technology-intensive intel companies. Intel's only heavyweight competitor so far is Motorola in 1980s. Just as the rise of the Roman Empire was achieved by defeating the original overlord of the bucket war, the rise of Intel was achieved by defeating the old semiconductor company Motorola. Motorola was founded in 1928. As early as the Second World War, it was a provider of wireless communication between the US military. Since 1960s, it has been ahead of the world in terms of communication and integrated circuits. Two years earlier than intel, Motorola launched a 16-bit microprocessor with a decimal computing performance of 5 times better than 8086 68000. The name 68000 is obtained based on the number of transistors it integrates into 68,000. While 8086 only has less than 30 thousand transistors. At that time, many workstations, including HP, sun, and Apollo, were all using motorola processors. In Intel's 80286 year (1982), Motorola launched a 80286, which was significantly better than 68010 in terms of performance and continued to serve as the processor of the current main workstation. It is said that intel has played a trick in the model to compete with Motorola. Intel's second-generation processor should have been named 80186, however, Intel left this number to an unimportant input/output processing chip, and jumped its series processor number to 80286, people who do not know the technology think that Intel's processor is a generation higher than Motorola's. In the 32-bit microprocessor competition, Motorola had the upper hand in terms of technology and release time, and its next 68020 was significantly better than Intel's 80386, except for being used on the main workstation, 68020 was also selected by Apple as the Macintosh processor.

At this time, intel received strong support from the outside. With the gradual popularization of ibm pc compatible machines, Intel, which is relatively backward in technology, has occupied more market share. Although Motorola later introduced 80486, which corresponds to Intel 68030, at this time, various workstation companies began to develop their own reduced-to-point instruction processor, motorola has only one Apple user left and it is difficult to compete with intel. A few years later, Motorola simply joined the ranks of world leaders in the competition for the Power PC. Ten years later, as Apple began to use intel's processor, Motorola completely launched the microcomputer processor market.

Motorola was not defeated in technology and funding. Before 1980s, Motorola was significantly stronger in terms of capital and technology than intel. For a long time, its processor is superior to Intel's similar products in terms of performance. The defeat of Motorola was first caused by Microsoft, that is, Intel had an unsigned Allied power with Microsoft. However, Motorola has made many mistakes in business, management, and market. If Motorola operates properly, it should be able to hold its workstation and Apple's market today with a streamlined Instruction Set Processor.

To analyze Motorola's defeat, we may wish to compare intel and Motorola. First, they are two companies of different ages. Although Motorola, headquartered in Central Illinois, is also a high-tech company that experienced the information revolution in 1980s, its operating status is still a traditional company of 50 or 60 s. Although Motorola has a good salary and benefits for its employees, the company and its employees are basically in the traditional employment relationship. There are many internal management levels in the company, most employees do not have much stock options. Therefore, the company's performance has little to do with the interests of its employees. Intel is a typical Silicon Valley company. Each employee has a much higher job intensity than Motorola, but each employee has more average stock options. Some of the houses in several good school districts in Silicon Valley have been bought by Intel's Early employees, and these houses cannot be bought for a lifetime by salary. Several years ago, the American History Channel commented on the Sino-Japanese War in the program. The American historian believes that this is a war between two armies of different ages. Although the weapons on both sides are similar, there will be no suspense in the results of the war, it is difficult for an army in the late stage of the authoritarian agricultural age to win an army in an industrialized country. The same is true of competition between Intel and Motorola.

Second, the two companies are far behind each other. Although a scholar and the best leader and manager in the computer age, Intel's CEO groove in the 80 or 90 s was named the world's best CEO several times. Motorola was founded by Galvin's brother. In 1960s, the company was transferred to its son, And in 80 or 90, it was transferred to its grandson. It was a typical family company. As the saying goes, there are only three generations of rich people. This is true for the Garvin family. The three generations are not the same. Although his grandson's Christopher Garvin was "elected" as CEO, if he was not surnamed Garvin, he would never be the CEO of Motorola, or even fail to enter the senior leadership of the industry.

In terms of business, semiconductors are only a Department of Motorola, while microprocessor is only a business of its semiconductor department, but it is all for intel. Therefore, Motorola's full withdrawal from the microprocessor market is only a loss of some territory, and once intel fails, it will face a disaster recovery. Generally speaking, Wall Street always wants a listed company to have as many as possible rather than a single source of income. Motorola does, it has developed in many fields such as computer processor, communication digital signal processor, walkie talkie, BP machine, mobile phone and TV receiver. As a result, it is difficult to expand in every field. Intel has been very focused on its personal computer processors. R & D of each generation of products is concentrated with a large amount of manpower and funds, and each time it is successful, it cannot fail. This is like a wire and a rope. A wire is easily broken by a rope. Therefore, dedicated intel eventually made the computer processor business very large and good, while Motorola, with its diversified business, finally lost to intel in addition to its microprocessor, and met Nokia on its mobile phone, the signal processor (DSP) was defeated by Texas Instruments ). Many people asked me if Yahoo could catch up with Google in the search field. I specifically replied-no, because Yahoo cannot focus on this field. Sometimes, a good company cannot do what Wall Street wants.

If time can go back, let Motorola and Intel change at that time, that is, ibm pc uses Motorola's processor, and the server manufacturer and apple are handed over to intel. Then, after 20 years of development, Motorola is also hard to become the leader in the semiconductor field because its internal problems cannot be solved.

3. Competition for Instruction Sets

Intel, with the help of Microsoft, won the first battle of Motorola in business. In the next ten years, it has fought the world again technically.

The structure of today's computer systems can be divided into complex commands (CISC) and simple commands (RISC) based on the instruction set. A computerProgramFinally, a series of commands can be run on the processor. The instruction sets of each processor are different. Some processors are designed to implement as many instructions as possible, including all the early IBM and Dec computers, and today's intel and AMD processors. The advantage of using a processor chip of a complex command system is that it can implement complicated commands, but there are two main problems. First, the design is complicated and the integration required to achieve the same performance is high; second, because the execution time of each command is not the same, it is difficult to run a flow task in each part of the processor, and the processor will wait unneeded. In addition, there is also a defect in the past, that is, the high power consumption caused by the high integration of complex command chips.

In 1980s, computer scientists proposed a processor design concept based on a simplified instruction set, he is now the principal of Stanford University, Professor John Hennessy of the American Emy of sciences, the College of Engineering and the three schools of arts and sciences, and academician David Patterson of the famous computer professor at Berkeley. The simplified command system only keeps few common commands and replaces a complex command with several simple commands. The design concept based on the simplified instruction set is a revolution in the history of computer development. It greatly simplifies the design of the computer processor. At the same time, because the processor of the simplified instruction set can ensure that the execution time of each instruction is the same, each part of the processor can flow well, and the processor speed can be faster than the processor based on complex commands in the same period. The streamlined Instruction Set Processor includes a lot of workstation processors and now the fastest Sony PS/3 game machine microprocessor PS/3-cell.

Although the processor of complex commands and streamlined commands has their own merits, the academic community almost thinks that the design of complex instruction sets is outdated, and the streamlined instruction sets are advanced. In particular, all U. S. university computer principles and computer system structures are textbooks written in conjunction with hennesis and Paterson. For a long time, the book mainly introduced the MIPs streamlined command chip designed by hennesis. At the same time, the papers on IEEE and ACM system structure mainly focus on simplified commands. Intel has not yet designed a chip for command streamlining at 8086. Otherwise, I think Intel is likely to adopt this technology rather than a complex command system. Once complex commands are taken, intel must continue to use complex command systems to be fully compatible with 8086. In the mid-to-late 1980s s, many processor with simplified instructions were made, including the MIPs designed by hennesis, which were later used for SGI workstation, also for the CPU design by Paterson, and later for IBM workstation. The speed of streamlining command chips was much faster than that of complex commands.

By the end of 1980s, Intel was faced with the option of continuing to design a chip compatible with the previous x86 version or transferring it to a streamlined command. Intel's market advantage will be no longer on the road to streamlined instructions; if it sticks to the path of complex instructions, it will have to go against the world's processor development trend. Intel is very rational about this issue. First of all, intel must maintain its leading position in the microprocessor market through x86 chips. However, if the processor with complex instructions develops, simplified commands represent the future development direction, and they cannot be left alone. Intel has launched a 80486 processor based on simplified instruction sets while introducing a 80860 processor based on a complex transitional instruction set. This product proved to be unsuccessful. Obviously, the market trend shows that the compatibility requirements of users are more important than performance. Therefore, after Intel launched its 80960 simplified command, it stopped its work and focused on the complex command series that are "technically backward. In the 1990s s, only one intel processor insisted on developing complex instruction sets against the entire processor industry.

it should be said that intel is not doing anything about streamlining command processors. It draws on the strengths of the extended and later processor designs, this improves the efficiency of the processor's internal pipeline. Because Intel's sales of each type of PC processor are more than the total sales of all of its workstation processors, it can invest more R & D funds and manpower in the development of each processor than in any kind of streamlined instruction processor. In this way, Intel has invested heavily, this ensures that the processor performance is faster than the simplified command. In the simplified command camp, in 1990s, five workstation manufacturers sun, SGI, IBM, Dec, and HP fought independently, each of which produced its own streamlined command processor, and Motorola's PowerPC for Apple, the six companies split up a market, and eventually no one was doing very well. Before and after 2000, various processors could not be processed, or all or part of them began to use Intel products. However, almost no one uses the MIPs processor, which was the earliest to streamline the command. Hennesis and Paterson, as two responsible scientists, add Intel processors to their textbooks to prevent college students from learning dinosaurs such as MIPS.

After ten years of hard work, Intel finally won the battle for a processor for a streamlined instruction set. It should be emphasized that Intel won this battle not by technology, but by the market. Intel is worthy of attention in many places. First, Intel adheres to the compatibility of its series of products, that is, to ensure that previous software programs will certainly run on new processors. After a long time, users have accumulated a lot of software running on Intel processors. Each time the processor is upgraded, the user's original software can be used, which is very convenient. Therefore, we do not want to easily replace the processors from other manufacturers, even if those processors are faster. Other processor manufacturers are not doing well with intel. They often restart every few years, causing users to be unable to use a lot of software and have to spend money to buy new ones. After a long time, the user will get bored. Second, Intel leverages the advantages of economies of scale and invests heavily in R & D, so that the industry's sophisticated Instruction Set Processors are updated from generation to generation. In early 1990s, Intel's X86 series and simplified Instruction Set Processor were slightly inferior to real number operations. However, Intel has been making unremitting efforts for more than a decade. Other vendors are far behind intel because their respective markets are not large enough. Rather than defeating other vendors, Intel is better at killing its competitors. Third, Intel has not rejected new technologies. It has also developed two good, streamlined instruction processors, but immediately stops them when it sees their future is poor. Fourth, Intel is lucky. It has no choice in the simplified command processor camp. In this battle, it seems that intel is the leader in many processor fields. However, because these companies are competitors in workstation, they naturally do not use their products. They also fight independently and split their servers to fight for a price war. Finally, sun and IBM have packed up several other workstation companies, but they cannot compete with intel. Now they have used Intel chips. Originally, Motorola was most likely to stand out from intel in the world of streamlined command processors, because it itself is not a workstation, and all workstation manufacturers originally used its 68000 series Processors, however, Motorola does not compete with each other. We have analyzed the cause.

4. Relationship between Intel and AMD

We mentioned previously that Motorola used the "Battle of Intel and Motorola" statement, because it is indeed a very dangerous war for intel, at that time, Motorola was superior in both technical and financial aspects. If intel goes wrong, it will not exist today. The battle between Intel and many simplified command processor companies can be said to be shocking, because intel is already internal and external. Today, competition between Intel and AMD in the market I personally think it is not a heavyweight competitor, so it is not a war. I think, if it is not the anti-monopoly law, Intel may have killed or acquired AMD. In addition, the relationship between Intel and AMD is basically a joint and struggle.

Unlike Intel's competitors, AMD has never made a different chip from intel, but is constantly introducing cheaper alternatives compatible with intel. AMD's practice is closely related to its genes. AMD should be Intel's family brother in terms of kinship, because it is also derived from xiantong semiconductor, also in Silicon Valley, only a few years later than intel, and also like intel, starting from semiconductor memory. Unlike other processor companies, the founder of AMD is from a sales background, while the founder of a general technology company is from a technical background. AMD's genes determine what it will do instead of what it will do, but market-oriented, and what the market needs. Soon after amd was created, it successfully unshaved an Intel eight-bit processor chip. In 1980s, Because IBM had to purchase more than two companies to compete, Intel had offered amd to sell chips to companies such as IBM for a long time.

in 1986, intel did not want amd to produce the first 80386 of its profits. It may want to exclusive 80386 of its profits, so it began to break the contract. AMD took out the previous contract request for arbitration, and the arbitration result was that AMD could produce 80386. Now, intel did not do it. He appealed to the California High Court. This lawsuit lasted for several years, but the court basically upheld the arbitration result. AMD simply clones Intel's processor chip. At that time, computer manufacturers, such as Compaq, began to purchase a small number of AMD chips in order to lower prices with intel. A few years later, Intel once again accused amd of stealing the MMX technology it bought for several hundred million dollars. AMD made concessions to reach a settlement. During the 1990s s, although Intel and AMD had a fight, they had the same benefits in opening up the x86 market against workstation chips with reduced instruction sets. Therefore, they are more dependent on the market than on competition.

two thousand years later, the US economy fell into a slump, and the workstation market for simplified instruction plummeted. Sun's shares fell by more than 90%. The processor market is dominated by Intel and AMD. AMD took the initiative to use it to develop a 64-bit processor advantage, first challenged intel in the high-end market, and won a lot of shares in the server market. In the past few years, Microsoft was unable to launch a new operating system, Vista, so individual users had no motivation to update their computers. At the same time, due to the development of the Internet, the network server market grew rapidly, demand for 64-bit high-end processor chips has increased significantly. In this way, AMD's performance has continuously increased over the past few years. It once occupied around 40% of the processor market and stirred up a price war with intel. AMD also sued intel for its monopoly around the world. By the beginning of last year, amd not only peaked in performance, but also gained some benefits in its anti-monopoly lawsuit against Intel. The European Union and other countries began to constrain intel. In this way, intel cannot underestimate AMD's brother. It decided to show amd some colors. In the next year, Intel's core-core dual-core processor has finally been available, and its performance is higher than that of AMD products. Intel has regained its leading position in products. At the same time, Intel has moved its production lines to lower-cost Russia and Arizona, which is much lower than Silicon Valley in the past few years. Then, Intel began to respond to the price war. As a result of the price war, Intel's profit margin was somewhat affected, but amd suffered a huge loss from profit. Intel has regained its initiative in the processor market. At present, both companies adopt a semiconductor technology of up to 65 nanometers. In the next few years, Intel will be significantly ahead of AMD in terms of the latest 45nm technology, and has begun to develop a more integrated 32nm chip, it will maintain an absolute advantage over AMD.

I think, in general, Intel does not want to completely kill AMD. Because AMD has more advantages than disadvantages for it. First, it avoids a lot of anti-monopoly troubles. Today, AMD has only 5% of Intel's shares, and the latter's cash on hand is enough to buy the former. However, intel cannot do this. Otherwise, there will be great anti-monopoly problems. Second, keeping AMD's opponent is good for Intel's own technological advances. Liu Zongyuan pointed out in his article "enemy ring" that "Qin has six countries, and Qin has a strong affinity ". This rule also applies to intel. Intel has increased the processor speed (if measured by decimal speed) By 1979 times since 0.25 million. Without many competitors, it cannot do this. Now its main competitor is AMD. From the perspective of motivating itself, it may be necessary to keep it. After all, AMD is technically not as technical as Motorola and IBM in the past. With such a joke circulating, Intel met AMD's peers one day and said, when will your new processor be made, we will have new things to do.

5. Difficult days

Peter, president of Google research institute and author of the "artificial intelligence" textbook in the United States. dr. norwger has a classic saying in the industry that when a company's market share exceeds 50%, there is no need to double the market share. The implication is that the company must explore new growth points. After 2000, Intel was positioned like this. Now, it has basically monopolized the general-purpose processor market, and how to develop it in the future must be considered.

Although intel still occupies only one small segment in the entire semiconductor industry, it is not necessary to enter many markets, especially low-end markets, such as the memory market where Intel does not, therefore, its growth space is not as big as imagined. Intel is specialized in processor and PC-related chip manufacturing, so it is easy to develop into these two markets. However, so far, it has not been very successful in the Development of chips outside the microprocessor. For example, in the past few years, it spent several hundred million yuan to develop the peripheral chip of the PC, and finally ended in failure. Now it has to use the chip set of Marvell ). In addition to computers, many electrical and mechanical products now require processors. For example, a high-grade Mercedes-Benz car contains hundreds of chips with computing functions, the demand for cell phones for processor chips is not to mention. Intel once entered the market for high-end mobile phone processors. However, due to the high development cost of Intel, this department has been losing money and had to sell it to Marvell last year. So far, Intel's efforts beyond the microprocessor have all failed.

Intel's business model has always been to rely on large investment to make money in large quantities. For the same generation of chips, Intel's sales volume is 10 times or more than Sun's. Therefore, it costs several times more than other companies to develop a chip. However, when a chip market is small, it is hard for intel to make a profit. Now, this is the biggest problem it faces.

The second thing Intel wants to do is to prevent the resurgence of streamlined Instruction Set Processor companies such as IBM. Although in the personal computer market, intel X86 series of processors are irreplaceable for a long time, because Microsoft safeguards it in the operating system. But it is not necessarily in the server market, because the main operating system of the server is open-source Linux, and Linux can run on any processor, therefore, as long as there is a processor with superior performance in all aspects than intel, customers who purchase servers will consider using non-Intel processors. In today's energy shortage, server manufacturers are not only concerned with the speed of processors, but also the speed of per unit energy consumption. Now, if a core processor is used around the clock, the power consumption for one year is equivalent to the price. Therefore, energy consumption must be considered in the future processor design. Although intel has begun to pay attention to this issue in the past two years, Intel's complex command chips are generally not as simple as the processor design for simplified commands, but are relatively difficult to achieve low energy consumption. In this way, the processor of the simplified instruction set may compete with intel in the server market, although this will not shake Intel's foundation. In my personal opinion, in addition to personal computers, the most important market in the future is the gaming machine market. Currently, game consoles have not just been designed to play games. They have become the entertainment center of every family. IBM and other companies are at least leading in this field. IBM has monopolized the processor market for Nintendo, Sony, and Microsoft. In fact, these simple instruction processor-based game consoles have surpassed intel processor-based PCs in terms of computing speed and graphics Functions. If the next technological revolution takes place in the living room of every family, IBM will no doubt be able to get the lead.

Although intel is dominant in the PC Processor market, its vision is not optimistic with the saturation of the PC market. To some extent, it is the biggest victim of the antimoore theorem, because the price of the processor is declining. At the same time, it is difficult to step up in the development of new markets, it is difficult to get rid of the "novege effect" shadow. Fortunately, Intel is also the direct beneficiary of Andy Bill's theorem. in the foreseeable future, its development will depend heavily on updates of software from companies such as Microsoft.

Conclusion

In the personal computer era, it is very easy to assemble or even manufacture a microcomputer. Even I have collected my PC and sold it. Therefore, over the past 20 years, there have been countless computer brands, from compatible machines with their own brands collected from Zhongguancun to so-called brand machines that occupy the vast majority of the world's markets, such as Dell, HP, and Lenovo. Although these computers are widely used in configuration and performance, they all use Microsoft's operating systems and Intel series Processors. From this point of view, there are only two leaders in the microcomputer era: Microsoft in software and Intel in hardware. Some people even call the PC industry the Intel/Microsoft system.

Intel's biggest contribution to the world is that it proves that processor companies can exist independently of computer machine companies. Before intel, all computer companies had to design their own processors, which made computers very expensive and unavailable. Intel constantly provides cheap and better processors for various users all over the world, so that personal computers can be popularized directly. Its large investment and large volume of practices have become a model of today's semiconductor industry. It is undoubtedly one of the most successful companies in the information revolution in the past two decades, but in the future, unless it can find new growth points, otherwise, it will enter the peaceful period of the PC age.