Evolutionary history of supercomputers in China

How did Chinese supercomputers grow from nothing, from mainframes to supercomputers? In March 1950, the Central People's Broadcasting Station broadcast a letter, in which it said, "Although Liangyuan is good, it is not a hometown for a long time, and I will come back when I return." The author of the letter is Hua Luogeng, a mathematician who just arrived in Beijing from the United States via Hong Kong.

In the letter, Hua Luogeng shouted out the deafening words: "Science has no borders, scientists have their own motherland."

"Liang Garden is good" comes from a sentimental sigh of Sima Xiangru in the Han Dynasty when he bid farewell to King Liang Xiao, showing Sima Xiangru's lofty aspirations in the Quartet.

At that time, many Chinese scientists went to Europe and the United States to escape the war, devoted themselves to scientific research, and received preferential treatment.

Hua Luogeng has been hired as a professor by the University of Illinois in the United States, and the salary is generous.

Under the call of New China, many scientists, including Hua Luogeng, resolutely chose to return to the motherland from overseas.

After returning, Hua Luogeng returned to Tsinghua Garden and served as the dean of the Mathematics Department of Tsinghua University.

In the summer of 1952, feeling that the United States was carrying out research on electronic tube computers, Hua Luogeng advocated and took the lead, and together with Min Naida, Xia Peisu, and Wang Chuanying, established China's first electronic computer research group in the Institute of Mathematics of the Chinese Academy of Sciences.

At that time, China's computer field was deserted, and the forerunners were holding torches high to explore the unknown.

Min Naida wrote China's first "conception and planning of electronic computer research", which opened the prelude to the development of electronic computers in my country.

In March 1956, in the "Outline of the Prospective Plan for the Development of Science and Technology 1956-1967" formulated by the People's Republic of China, the computer became one of the "four emergency measures" for the development of science and technology.

At the same time, Hua Luogeng was appointed as the director of the Preparatory Committee of the Institute of Computing Technology, Chinese Academy of Sciences, to develop a general-purpose digital electronic computer, and at the same time sent someone to learn from the "big brother" the Soviet Union.

In 1958, under the principle of "based on a domestic imitation first, and then self-designed", China successfully developed 103 machines that imitated the Soviet M-3 mainframe computer, with a computing speed of 1,500 times per second, achieving a leap from zero to one.

Although China's computer research and development are only 12 years away from the invention of the world's first computing ENIAC by the United States, there is already a considerable distance between them in terms of talents and equipment.

The United States has carried out computer research since World War II. There are hundreds of thousands of experienced scientific researchers from the military to the private sector, and the funds are sufficient.

At this time in China, let alone computer research, many people have not even figured out what a computer is, let alone seen it with their own eyes, and funding is even more stretched.

At that time, with the help of Bell Labs, IBM in the United States successfully developed the RCA501 computer with a transistor structure.

Due to the use of transistor logic elements and fast magnetic core memory, the calculation speed of the RCA501 computer has increased from 5,000 times per second to hundreds of thousands of times, and the storage capacity of the main memory has also increased from several thousand KB to more than 100,000 KB. A simple operating system using a high-level language and its compiler.

Compared with traditional electronic tube computers, transistors can not only realize the functions of electronic tubes but also have the advantages of small size, lightweight, high efficiency, low power consumption, etc., so that the size of the computer has been reduced from occupying a room in the past to only the size of a desk.

Faced with the shortage of talent, the Preparatory Committee of the China Institute of Computing Technology has cooperated with many famous universities in China to hold four consecutive computer training courses and has trained more than 700 scientific researchers in computer fields, laying a solid foundation for the talents of China's computer industry. Base.

In the wild era, this small team lit up China's computer industry like a faint fluorescent light on a long night.

After the Sino-Soviet "honeymoon" ended, the Soviet Union withdrew all aid to China, including computer experts and equipment.

China's computer scientists have not stopped, relying on collective wisdom and dedication to overcome difficulties in turbulent times.

In 1960, Xia Peisu led the team to successfully develop China's first self-designed small general-purpose electronic digital computer 107 and delivered it to the University of Science and Technology of China in Beijing at the time. This was also the first scientific research computer in a university in China.

In 1964, my country's first self-designed large-scale general-purpose digital tube computer 119 was launched; in the same year, Fudan developed a 602-type electronic digital computer programmed in machine language.

In June 1965, the Institute of Computing Technology of the Chinese Academy of Sciences successfully developed the first transistor computer 109 B, with an operating speed of 90,000 times/s for fixed-point operations and 60,000 times/s for floating-point operations.

The subsequent 109C aircraft played an important role in the trial production of the "two bombs" and was known as the "merit aircraft".

Throughout the 1960s, domestic computer research and development were carried out around major national defense projects, only pursuing continuous improvement in computing speed, and not much consideration was given to the overall performance and popularity of computers. There is no concept of mass production.

In the 1970s, developed countries such as the United States and the West have developed miniaturized integrated circuit general-purpose digital electronic computers, which have been widely used in the private sector.

Therefore, in January 1973, the Fourth Ministry of Machinery Industry held the "First Professional Conference on Electronic Computers (7301 Conference)" in Beijing, making it clear that in the future "we must give up the technical policy of simply pursuing the improvement of computing speed, and determine the policy of developing a series of machines."

After this meeting, China began the rudimentary development of the computer industry in the following ten years.

Amid the red flags and the clamor of gongs and drums, Chinese computer researchers are stepping up research and development while outlining the computer models needed for economic construction.

In August 1973, my country's first mega-integrated large-scale computer 150 was born. The main memory of this computer is 130K. Equipped with multiple programs and operating systems, it can perform 1 million operations per second.

Since then, China's large-scale computers have gradually shifted to the level of economic construction, taking on new missions in China's oil exploration, weather forecasting, scientific computing and other fields.

The DJS-131 computer produced in 1975 is a desktop computer with a memory of 4-32K (expandable to 64K) and an operation speed of 500,000 times per second.

A total of 334 DJS-131 minicomputers were produced, but it was the most widely used and most stable domestic digital electronic computer in China at that time. Research, transportation, industry and national defense construction, and other fields.

At the same time, the North China Institute of Computing Technology organized 57 units across the country to jointly develop and design DJS-200 series computers and DJS-180 super minicomputers, which opened the prelude to the localization of microcomputers.

The NCI-2780 super minicomputer, TJ-2000 series machine, AP array processor, and other products appeared one after another, which gradually made China's computer industry embark on the road of serialized mass production.

Always adhering to "walking on two legs", China's computer industry continues to maintain gratifying achievements on the road of miniaturization and large-scale.

In November 1976, the Institute of Computing Technology of the Chinese Academy of Sciences successfully developed a large-scale general-purpose integrated circuit general-purpose digital electronic computer 013 with a computing speed of 10 million times per second, which was then widely used in aerospace, national defense construction, and other fields.

Those who participated in the research and development of these projects were all the students who came out of the training courses at the China Institute of Computing Technology.

The once-fluttering fire has finally become a scientific pedigree, spreading its branches in various application fields of China's computer, and starting the inheritance.

Through independent research and development, China's computer industry has completed the transition from the second generation to the third generation, especially through the application to promote the development of computer research and development, the combination of production and use, and promote the localization of microcomputers.

Although China has been persistently conducting research and development in the field of large computers, there is still a considerable technological gap with the United States.

The United States has developed the third-generation computer with integrated circuit (IC) as the main body since the early 1960s and began the process of commercialization.

Among them, the IBM 360 system mainframe computer released by IBM can run many different programs at the same time and extends to the field of word processing and graphics processing, and the mainframe computer is many times ahead of China.

Take the more advanced 104 machines in China at that time as an example. This large computer, which later participated in several large-scale national defense scientific research projects, could complete 10,000 floating-point operations per second; while the 7030 machine designed by IBM for the US military at the same time, floating-point The computing performance is 600,000 operations per second.

The performance of floating-point operations determines the efficiency of scientific research and the accuracy of the calculation. China has tried to spend more than 7 million US dollars to buy it, but due to the special historical environment, it has not been able to do so.

Even in the field of minicomputers, the United States has taken a big step forward.

In the 1970s, the United States began to try to use Unix and C language for programming through DEC's minicomputer PDP-11, to meet the needs of various scenarios, and truly let the computer enter the stage of "serving the people".

At that time, China's large-scale computers still difficulty meeting various construction needs in terms of performance and output and were still struggling.

In 1976, Cray Corporation of the United States launched the world's first supercomputer with a computing speed of 250 million times per second.

"Mainframe computer" and "supercomputer" may seem similar, but there is a world of difference between the two.

To put it simply, large computers use special instruction systems and operating systems. Since they can only perform non-numerical calculations (data processing), their application scope is generally limited to the commercial field; while supercomputers use general-purpose processors and UNIX or UNIX-like operating systems, which are good at Numerical computing (scientific computing) can be widely used in more sophisticated and complex scenarios such as national defense, aerospace, meteorology, and industry.

Supercomputers are true "the most important weapon of the country", which brings together tens of millions of processors and can perform massive data high-speed operations.

Facing the reality gap, the Chinese government decided to purchase equipment to learn and master the policy of supercomputer cutting-edge technology.

After China and the U.S. resumed contact, the U.S. exported two supercomputers to China in order to express friendship, but Chinese computer researchers experienced a humiliating "glass room" history.

At the end of 1976, outgoing US President Ford signed an export agreement for two Cyber172 supercomputers to China.

However, the US has manipulated this computer, and the computing performance is far from the actual level, and the US also requires that the computer can only be used for geological exploration and cannot be used in other ways.

What annoys Chinese scientists the most and is helpless is that the computer is set up in a special glass machine room, but the keys are managed by the United States. Before each use, it must be approved by the Americans.

After the operation is completed, the US will immediately block the glass room, and the operation log will also be regularly submitted to the US government for review.

After the introduction of the US Cyber172, China purchased three Hitachi M-series supercomputers from Japan for weather forecasting.

But the U.S. made a fool out of it, not only requiring the Japanese side to significantly reduce the performance of the version exported to China but also requiring the establishment of an on-duty personnel and monitoring log review system in accordance with the practice of Cyber172.

This situation continued until the early 1980s when China purchased various types of large-scale and supercomputer equipment from the United States, and key technologies were always tightly sealed by the other side.

For Chinese computer scientists, this is an unforgettable past.

Yang Xuejun, an academician of the Chinese Academy of Sciences and former president of the National Defense University of Science and Technology, said about this history: "This is the eternal pain in the hearts of Chinese scientific researchers, just like farmers have no food at home, and mothers have no milk to feed their children..."

Stimulated by repeated "stuck necks", Chinese computer researchers realized that trying to achieve self-improvement by purchasing foreign advanced equipment is an unrealistic fantasy, and self-reliance is the only way out.

March 1978: Deng Xiaoping listened to the report on computer development, and made it clear that the National Defense Science, Technology and Industry Committee system will undertake the development of the first billion-time supercomputer, saying that "China must carry out four modernizations without supercomputers (supercomputers)."

In May of the same year, at the demonstration meeting of China's supercomputer plan, the project was named "785 supercomputers", and then General Zhang Aiping, director of the National Defense Science and Technology Commission, named it "Galaxy".

The Galaxy computer has a planned performance of 100 million floating-point operations per second, which is 100 times faster than the most advanced computer ever developed in China.

The director of the Computer Research Institute of Changsha Institute of Technology (National University of Defense Technology) who accepted the task, Ci Yungui, who is over sixty years old, issued a military order, "Even if you risk my old life, we must also create our own supercomputer!"

Ci Yungui is a generational master in the Chinese computer industry. From tube computers to transistor computers to integrated circuit computers, he is a major player in every upgrade of Chinese computers.

The 785 supercomputer project team also stated: 100 million times per second, once a lot! Six years, not a day! Budget, no more than a cent!

There are grievances that can be talked about, and difficulties that must be overcome. After all, the rise of the scientific and technological strength of a great country will depend on the unremitting efforts of a generation.

The scientific research team first referred to the overall architecture of Cray-1, the most advanced supercomputer in the United States, and then began to divide and collaborate. If there are supporting equipment for direct use in China, and independent research and development cannot be achieved, they are directly imported from countries with loose export conditions for parts and components.

But the team has a basic principle: the equipment can be used in other countries, and the technology must be it is own.

In order to achieve this goal, the researchers checked one by one in the test tapes piled up in small mountains; in order to ensure the stability and reliability of the machine, 25,000 winding wires, 120,000 winding points, and more than 2 million solder joints were carefully checked. .

In the end, the Ciyungui team creatively proposed the "dual vector array" structure, which greatly improved the computing speed of the machine and realized the planned development task one year ahead of schedule.

On December 26, 1983, China's first billion-time supercomputer "Galaxy No. 1" passed the national technical appraisal and was born.

"Galaxy-1" is an important milestone in the development of China's high-speed computer, marking that China has become the third country after the United States and Japan to independently design and manufacture supercomputers.

After the celebration party, each key R&D staff was given a bonus of 500 yuan, but none of them asked for it, and all of them donated to the computer project they were working on.

It is a pity that during the next 20 years of research and development of the "Galaxy" series, many people in the team passed away due to illness due to high work intensity and insufficient nutrition and medical conditions.

These researchers know that research funding is limited and their salaries are meager. In order not to cause trouble for research and development, they have chosen to conceal their illnesses, and they have been procrastinating.

They are 49-year-old Professor Zhong Shixi, 43-year-old Associate Professor Jian Xianfu, 41-year-old Associate Researcher Wang Yumin, 40-year-old Lecturer Zhang Shusheng, and 35-year-old Assistant Researcher Yu Wulong, they are worthy of our memory!

Among the relics of these researchers is the book "The Third Wave" by American author Alvin Toffler, the last sentence of which was redrawn by many people with red and blue pencils "like the ancestors of the revolution" Likewise, our mission is destined to create the future."

Time bows and salutes here, the world is the most important weapon, and the country is unparalleled.

In the same year of the appearance of "Galaxy No. 1", the "Wubi Font" of the 26-key scheme was born. This was the first version of the Wubi font input method that became famous at home and abroad. The inventor Wang Yongmin was called the modern "Bi Font". Lift".

"Wubi Input" is matched with the domestic D-2000 Chinese intelligent terminal and ZD-1110 character display terminal. The ancient characters and the computer system representing modern technology meet here. Those beating characters collided into sentences and spliced ​​into words, interpreting the cultural world belonging to the Chinese on the screen.

This year, China's first large-scale vector computer system, the 757 machines, was successfully developed, with vector operations reaching 10 million operations per second and scalar operations reaching 2.8 million operations per second.

Chinese scientific research institutions have also used the microcomputer Great Wall 100 (DJS-0520 microcomputer) developed by the Sixth Institute of the Ministry of Electronics. This machine has initially acquired the main features of personal computers.

Outside the window is an information-based world, and the "third wave" has swept across, opening a new chapter in human society.

Under the background that IBM's PC-compatible machines have become the world's mainstream personal computers, the Chinese computer industry proposed to "do as IBM's PCs".

It's just that this kind of "doing it" actually has no real objects and no drawings, and it is entirely up to the computer manufacturers to explore by themselves. Most of the computers that were finally "touched" were nondescript, and neither the performance nor the operating system could be mass-produced.

On February 16, 1984, Deng Xiaoping touched the head of a young computer contestant at the National Ten-Year Science and Technology Achievement Exhibition held in Shanghai, and said firmly, "Computer popularization must start with a baby."

The old man's advice made computer courses such as a blowout. Many schools have opened computer courses, ranging from simple typing exercises to advanced Basic programming, which has ignited the enthusiasm of Chinese people to learn computers.

People flick their fingers on the keyboard like flying as if confiding the urgent voice of an era of forging ahead.

In 1985, the Chinese people realized the need for complete Chinese information processing on the computer through the domestic microcomputer Great Wall 0520CH loaded with the Chinese operating system.

In the same year, the computer company that sold Lenovo Hanka was renamed Lenovo.

Driven by the Great Wall, Lenovo, and other brands, a large number of computer manufacturing enterprises have emerged in China, such as Stone, Founder, Tongchuang, Shida, etc., which have become the leaders in driving the development of China's computer industry, allowing China's computer industry to enter the third generation. to the fourth generation.

It was a rich history, but it was followed by a lamentable period of stagnation in China's computer industry.

As Sino-US relations become increasingly hot, many advanced technologies can be introduced smoothly. Therefore, the development model of my country's computer and semiconductor electronics industry has changed from "innovation-based, introduction-assisted" an emphasis on basic research to pure introduction, giving up the catch-up plan for semiconductor general circuit research and development.

The integrated circuit and semiconductor industries, which were originally not far behind the world, were suspended due to lack of funding, a large number of scientific research teams were disbanded, and some researchers were even appointed as computer room administrators, at a loss.

In fact, my country could already imitate CPU chips such as 6800 and 8080 at that time, and it was not inferior to many developed countries in technology.

However, due to the shrinking of project funds and the missed research and development direction, the CPU research could not be carried out independently and continued, thus laying a hidden danger for the computer chip industry in the future.

Fortunately, some experts in the Chinese scientific community have realized the gap between themselves and the international frontier technology, and they continue to call for it, which has enabled China's supercomputer engineering to catch up at a faster rate.

In March 1986, four scientists, Wang Daheng, Wang Ganchang, Yang Jiachi, and Chen Fangyun, based on the powerful scientific and technological strength behind the Star Wars program proposed by the United States, put forward a proposal to catch up with the world's high-tech.

Deng Xiaoping expressed support for the proposal, and soon, the "High-Tech Research and Development Plan" was released. Because the four scientists submitted the letter and Comrade Deng Xiaoping's approval was in March 1986, the plan was also called the "863 Plan".

A year later, at the Beijing Computer Application Technology Research Institute, the Chinese sent the first e-mail: "Across the Great Wall we can reach every corner of the world."

Since then, China and the world have been closely linked.

In the 863 plan, "intelligent computer" is listed as a separate subject project (ie 306 subject), and the project is undertaken by the Institute of Computing Technology, Chinese Academy of Sciences.

The Institute of Computing Technology of the Chinese Academy of Sciences is the cradle of China's computer development. The first general-purpose digital electronic computer and the first general-purpose CPU chip were born here. It is important behind the success of the "two bombs and one satellite".

At that time, the world had just set off a wave of artificial intelligence, and Japan's "fifth generation machine" developed rapidly in this field. Therefore, the "306 theme" followed the strategy of Japan's "fifth generation machine" and started the research on intelligent computers.

During the research and development process, experts found that the main goal of Japan's "fifth generation machine" is to achieve natural language translation, but the application market is narrow and the maintenance cost is high. At the same time, the US research and development objects are personal computers, supercomputers, and the Internet, and the prospects seem to be broader.

Time does not wait, the expert group decisively adjusted the research direction from "intelligent computer" to "high-performance parallel computer system", which gave birth to the accelerated development of China's supercomputer series.

In charge of the intelligent computer project, Professor Li Guojie, who just returned from studying in the United States, realized that at that time, domestic computer research and development was almost "closed door", and the research and development time was extremely long.

Therefore, Li Guojie sent a team to the Silicon Valley in the United States to "jump in the queue" to absorb foreign advanced technology and update it into the project synchronously. Moreover, he also developed and designed the domestic operating system SNIX on the basis of the original UNIX source code.

In October 1993, the peak computing speed reached 640 million times per second, and the "Dawning No. 1" supercomputer with a fully symmetric shared storage multiprocessing structure was born. The whole is similar to Intel's 1990 architecture and technology.

The subsequent Dawning 1000 supercomputer also started market operation and entered the ranks of enterprise services.

In 1997, with the establishment of the Dawning Tianchao 1000A in Liaohe Oilfield, China's supercomputer completely broke the market monopoly of imported products and washed away the stigma of the "glass room" in the former supercomputing field.

During this period, with the continuous emergence of domestic computer brands, the price of a personal computer that used to cost tens of thousands of yuan has dropped to less than 10,000 yuan.

In the 20 years since the new millennium, with the recovery of national strength and the layout of scientific research, China's supercomputer research institutions have been chasing after each other, and the development is in full swing.

In 2008, the Dawning 5000 supercomputer of the Chinese Academy of Sciences was successfully developed, with a computing speed of over one trillion times.

Just one year later, the "Tianhe-1" petascale supercomputer at the National Defense University of Science and Technology appeared, and China became the second country in the world to successfully develop a petascale supercomputer after the United States.

In June 2010, the "Nebula" petaflop computer of the Chinese Academy of Sciences won the second place in the 35th supercomputer TOP500 list, and successfully entered the top three ranks of the world's supercomputers.

Half a year later, the "Tianhe-1A" petascale supercomputer of the National Defense University of Science and Technology directly won first place in the ranking.

At the same time, the petascale supercomputer "Shenwei Blu-ray" of Jiangnan Computer Research Institute took the lead in completing the localization of the CPU.

On the Chinese supercomputing track, Dawning, Tianhe, and Shenwei have become the dazzling "Three Musketeers" of the high-performance computing special topic.

With the continuous deepening of supercomputing research, China has successively built seven institutions including the National Supercomputing Center in Jinan, Shenzhen Center, Wuxi Center, and Zhengzhou Center.

During this period, innovations such as cloud computing, big data, and artificial intelligence based on supercomputing architectures have also continued to stage new chapters.

In 2018, Dawning, Tianhe, and Shenwei have entered the E-level (10 billion operations per second) supercomputing research and development in the field of supercomputer competition, and gradually realize the national production of CPU and accelerator.

In 2021, in the fifty-eighth global supercomputer TOP500 list, 173 Chinese supercomputers will enter the list, accounting for 34.6%. The second place in the United States is 149 units, accounting for 29.8%.

However, although the number of supercomputers in China exceeds that of the United States, there is still a certain gap between China and the United States and other countries in terms of comprehensive computing power.

In the global supercomputing TOP500 list in the first half of 2022, China's Sunway TaihuLight ranked sixth, already the best among the "Three Musketeers".

However, there is still a gap of about 10 times in computing power from the world's first E-class supercomputer, Frontier of Oak Ridge National Laboratory in the United States.

It can be said that in the face of China's supercomputing system, the changes have never stopped.

In the 1960s, Gordon Moore, one of the founders of Intel, predicted that the number of transistors on an integrated circuit would steadily double every 18 months in the future, and maintain this momentum for decades to come.

Moore's Law has since become the law of the times, and it has become an evolutionary metaphor for the unknown world.

China's supercomputers have gone from barren to fertile fields, and they have reached the middle stage in the chess game of the ages. They are becoming clearer and clearer in the changing times, and they have gathered majestic power.

Post a Comment