For a man who spent 17 years on the chip of CPU, Hu Weiwu looked a bit tired. It was a clear morning in late March, and the reporter sat opposite him in the office near the North six ring in the northwest of Beijing. Hu Weiwu, 50, did not hide his admiration for Mao Zedong. On the wall of the company conference room hung longson research team armed with Mao Zedong thought. Hu Weiwus tone is not high, a straight Chinese tunic, with a badge of Chairman Mao on his chest and a pair of lightweight cloth shoes. Before the interview, he just sent away a group of guests, talking about the project of university cooperation. Such temperament concealed his strong belief and his bet on this belief. In the past 10 years, in the field of CPU, this has been considered to be in the semiconductor system, although a company that has not yet been able to carry the banner, but a 60 - point company in the field of computing chips. A senior analyst in the electronics industry said, if there are 60 enterprises in the domestic chip industry today, they will not be caught off guard. He said that because of the current domestic communication chip industry, there was no big business or even a passing line of 60 points, which led to the seven inch hit by the US sanctions. The analyst said this time is the US ban on ZTE chips. Longxin Zhongke, the chip company that is considered to be able to achieve 60 points, has worked hard and has died ninety-nine lives. Hu Weiwu said. Its a terrible bet to start a business in the chip industry. You know, most investors have been avoiding these areas just like mice and cats. This is essentially a very traditional manufacturing industry with a very low profit margin. A few years ago, when the semiconductor fund was set up on a large scale, semiconductor experts from Tsinghua University once told this newspaper. The era of the project group Walking into the building in the middle section of Longxin, the broadcast of Confucianism was read from the ear, with red slogans on the wall, the words of Hu Weiwus Chairman Maos quotations. These words were also printed on the red notebook of Godson. The cover of the notebook is Maos serving the people. Hu Weiwu is the same age as Intel (Intel). Intel has long been the top of the worlds largest semiconductor companies. In the second year of Hu Weiwus birth, in addition to Intel, the largest x86 architecture microprocessor supplier AMD was founded in San Francisco. Dating back to 2001, the Godson research group was formally established. In this year, the first integrated circuit appeared in the Chinese governments work report. This year, the worlds first universal microprocessor developed by Intel has been in existence for 30 years. A year ago, Intel, the worlds largest semiconductor company, released its four generation of Pentium processors, with a frequency of 1.3GHz. In the next two years, we will not start the general operating system (boot). On the way back from the alma maters University of Science & Technology China, Hu Weiwu, 32, called Tang Zhimin, the director of the computing architecture room, and began to take part in the CPU project. In the original office of the Godson research group, four upper and lower berths were arranged neatly, with simple sheets and bedding. The core members of the team are called Hu Weiwu. At that time, the dog was left behind was ruthless than the skin of a week, and sometimes even the members who had returned home were called to the office in the middle of the night. However, with the rapid expansion of Chinas chip deficit, the Chinese government has invested huge resources in trying to ripen local industries. In 2006, the Han core fraud incident was exposed. The Han core incident, known as the first technology fraud case of China in twenty-first Century, made people begin to rethink Chinese chips. Prior to that, in 2004, Jiangsu textile enterprise Meng Lan group and the Institute of computing technology of the Chinese Academy of Sciences jointly established the dream LAN core industrialization base. A year later, the laptop with Godson 2 came out. Ms. Qian Yue Bao has this feeling. At that time, she might have talked with Mr. Hu very much. 14 years later, Yang Xu, vice president of Godson Zhongke, recalled the scene of the year. Qian Yuebao is chairman, President and general manager of Jiangsu dream Lan group. The crisis of Han core has a great impact on us, and it is widely cursed on the Internet. Yang Xu said. Dragon core Han core is the same way, Han core is dragon core, dragon core is a pure trading company, and Meng Lan group is also involved. It is no wonder that the earliest dragon core industrialization partner will be a township enterprise in Jiangsu as a mattress mattress. In the same year, MIPS, the semiconductor design company of the United States, began to put pressure on Godson on issues such as patents. The Godsons command set 95% is similar to MIPS, which is plagiarism. At this point, Godson has developed its own EDA R & D platform, but in order to run Linux system, it must also be compatible with MIPS and other instruction set. In fact, the instruction set is equivalent to a language, for example, I invented Chinese, and you cant speak in Chinese, but you cant speak in Chinese, but I can authorize you to speak Chinese and then how you want to use it. This is not the same as some organizations buy chip design IP. IP is a novel written in Chinese, and the instruction set is authorized to create in Chinese. Lin Xiao agrees with this statement. Lin Xiao is a senior research fellow in the Institute of automation, Chinese Academy of Sciences. However, in the system of the Chinese Academy of Sciences, Godson is still a member of the Academy. At that time, multi-core CPU became a hot research direction of international academia. The CPU led by academics led to the relaxation of the single core performance during the 11th Five-Year period, but instead of being multi core. In Godsons commitment to the nuclear high base project, the Godson hastily launched the development of the 8 core chip, and even planned 16 core chips. At that time, the design level of the Godson CPU was more than 1/2 of that of Intel. In the past five or six years, the longcore processor has achieved 8 cores, but the single core level has only increased by about 50%. In contrast, the mainstream product is still dual core and 4 core, but the single core performance has increased by 5-10 times. The core is much more than Intel, but the single core performance is not enough, and many applications can not satisfy it. Out of the system A company named Beijing Tiantong Xinyuan Technology Co., Ltd. was founded. Shareholders include Hu Weiwu, Wang Jian, Gao Xiang, Yang Xu, Fan Baoxia and others. Tiantong is the Chinese translation of Godson. The company invested 5 million 500 thousand, accounting for 25.21% of Godsons share. When Longxin is established, the shareholders of the private background require the backbone to dry out. After all, the establishment of a business unit, how do you know that you are solid on the ground to do the dragon core? Financing is also the same. At that time there was no demand for autonomy, who voted for domestic CPU, who is fool, no one is willing to vote. Early investment in Godson has Beijings capital. Hu Weiwu said, once the leader of the city of Beijing went to Longxin investigation, he emphasized two sentences. The first sentence: CPU is the steel of the information industry. The role of CPU in the information industry is similar to that of the iron and steel industry; secondly, the level of our present CPU is similar to the steam car. We can not make BMW Benz, but already we have made it. It was made up of Santana and Charlie. The leaders of the city of Beijing were very supportive. Investors reluctantly, they feel that Godson will pay for it. They are all X86 things, and whom do you sell them to? At that time, Beijing industrial investment company, as a political investment, had already made good preparations for losing money. In April 2010, the Chinese Academy of Sciences and Beijing jointly led the investment into the stock market. Longxin completed the reform of the stock ownership structure. Longxin technology limited company started, formally moved to the industry, and the dragon core group also launched a transformation process to the company. In a sense, the development process of Godson 3B is the transformation process of this team twists and turns. The research and development of Godson 3B has been supported by the nuclear high base special project. This project is also a pilot project for the departments concerned to launch their own hardware and software applications. Nuclear high base is the abbreviation of core electronic devices, high-end general purpose chips and basic software products. It is one of the 16 major scientific and technological projects in the national medium and long term science and technology development program (2006-2020 years) issued by the State Council in 2006 with the manned spaceflight and the lunar exploration project. At the beginning of the company, the dragon core team has transformed from the project group to the company, but the R & D of Longxin 3B has also maintained the college orientation, a single indicator of excessive pursuit of multi core and floating point peak performance, and the general processing performance is insufficient. This direction has had serious consequences. Due to the lack of general processing performance, Godson 3B can not meet the urgent needs of independent information market at that time. Since 2013, nuclear high base has basically abandoned the CPU independent research and development route. At this time, the Godson CPU, which was mainly supported by national support, has been in a great predicament. After adjustment in the nuclear high base direction, at that time, there would be only a few millions of funds on the books, and the capital chain would soon be broken. Yang Xu said. In 2013, Godson moved from the original white house to the Zhongguancun environmental science and technology demonstration park in the northwest corner. Long distance relocation, plus the core salary of the core members at that time, was only 1.2 million, and some enterprises dug up 5-10 times the annual salary. Finally, members stayed in godson, this is my pride. Hu Weiwu said that after the company moved, many employees moved their homes to the demonstration garden. The independent research and development route has taken a detour, many people have doubts about independent research and development routes, and feel this road is blocked. At this point, nuclear high base has more support for the introduction of technology. During the period, Shanghai trillion core has undertaken the nuclear high base 01 special project, has obtained the huge fund. The subsidy is up to 5 billion 600 million yuan in the first phase of the project. The strength of the countrys determination and support can be seen. As a result of the new round of cooperation, foreign capital began to seek a joint venture with domestic enterprises or open the way of technical authorization. A bad way is to shake the original products of foreign capital into independent products. Under the stimulus of huge amount of money, it is hard to avoid the mud. According to the research staff in the field of chip design, for a project with a large number of national funds, we have a review of its own, the CPU running, there is a lot of information on the interface, there is a piece of information in it, that is, its research and development teams information, he hit XX hidden. Later, we recompiled the software, so that we could not hide it. The team shown in that code is the US team. Reporters have consulted the digital operators on this matter, but they can not confirm the authenticity from other sources. There are also a lot of foreign capital to find dragon core hope to cooperate, but Hu Weiwu refused, only build independent and controllable ecology, is the way out of domestic CPU, Hu Weiwu said, this is the ultimate goal of the dragon core. If we want to map out such an ecosystem in the civilian consumption market, it will take a long time. Zheng Zhenxiang, chief analyst of Zhongtai securities electronics industry, said he looked better at the establishment of an independent and controllable ecosystem in the party and army. If we want to build such an ecosystem in the field of party, government and army, it is obvious that the intervention of foreign capital is out of date. Obviously, at some point, Hu Weiwu will be very angry. In April 19th, Hu Weiwu said at the key information infrastructure independent security innovation forum held in the Wanshou Hotel, those who do not care for the interests of the national security for the interests of the small groups, those who do all the best to degrade the independent CPU, and shake up the banner of the comprador CPU route will eventually have to lift the stone to smash their feet. It is a disgrace for the development of Chinas information industry. From the end of 2014 to the beginning of 2016, about one and a half years, the performance of the autonomous CPU was criticized. It was required to replace the dragon core in the existing autonomous informatization pilot. It did change some of it, and caused great pressure to the dragon core. Sparks of fire When Godson entered the dilemma, the Chinese Academy of Sciences gave a sum of money. In May 2013, Godson also adjusted the R & D route of CPU. The direction of adjustment is to integrate CPU with specific applications, including aerospace, oil and flow meters. The special chip industry chain is short, easy to form technical advantages and quickly form sales. It is these products that are starting to form the sales revenue of Godson. At this point, the Godson team has launched a new round of financing. In 2014, Longxin gained the investment of CDH capital. CDH is one of the largest asset management institutions in China. Hu Weiwu said that the purpose of Ding Huis investment in Godson is to make money, he feels that Godson can make money. Godsons angel wheel was voted by the central government. After more than 10 years, social capital came. The reporter failed to get a response from CDH capital. Longxin company information said, from the second half of 2014, the role of the combination of dragon core R & D and market began to show, the sales revenue of Longxin company in 2014 was 51% higher than that in 2013; in 2015, it increased by 57% on the basis of 2014. In 2015, Godsons sales revenue exceeded 100 million yuan and was profitable. After the entry of CDH, the Godson 3A3000 has been developed successfully, which is of great significance to Godson. The performance of CPU is enough for transaction processing represented by Party and government offices. In October 24, 2017, Hu Weiwu said that the dragon core has now been used in more than a dozen national weights, including the Beidou satellite, and used in information systems such as functional offices. But, I cant sell it. It is still the biggest problem facing dragon core. The root cause of the problem has long been a consensus in the industry, that is, failure to build an autonomous and controllable ecosystem. In the civil market, the CPU of the server and PC market has long been the world of Intel and AMD. Intel is the whole business model barrier that covers intellectual property, technology accumulation, scale cost and software ecology. And this barrier has never declined. In Lin Xiaos view, such an ecosystem is difficult to build. The most intuitive understanding of the ecological problem is that 1 people are invested in the chip field, and the underlying firmware developed on the chip needs 6 people, and the underlying firmware does the corresponding operating system and ecological development, which requires 36 people. In this view, Lin Xiao expressed its identity. The software ecosystem is struggling, but Hu Weiwu insists that if we want to get recognition, we must adhere to the road of marketization. We do business, it starts with the negative. Hu Weiwu said that the dragons play was based on the party, government and army and expanded to the civilian market. According to Hu Weiwus practice of encircling the cities in the countryside, they start with small and medium-sized enterprises, and then cooperate with some universities to set up a pilot project. We must establish the position of market personnel above that of researchers. Hu Weiwu said. The world chip giant Intels research and development budget is calculated at tens of billions of dollars per year, while Godson has only a billion people and tens of millions of dollars. Hu Weiwu said Chairman Mao had told us that spark can start a prairie fire. Its hard to build an independent ecosystem, but its also the brightest way out. Hu Weiwu, 50, looked at the building of HUAWEI Research Institute outside the window and sipped a cup of tea.