Li Xuan dropped a bomb and left, leaving CUHK students chattering and discussing. Supercomputers are known as a bright pearl in the computer industry, and have been regarded as an important symbol to measure a country's scientific and technological development level and comprehensive national strength for a long time.

Later, it took 30 years and tens of billions of dollars to finally reach the world's fastest computing speed in the latest "Tianhe-2" supercomputer. This has also been jokingly called a stack chip game by countless netizens, and has been sprayed as a representative of high speed and low efficiency.

There is no doubt that supercomputers can play a role in the field of scientific research, especially in the field of defense industry, and it can be said that it is one of the standard configurations of major countries. Now China should be in full swing to develop China's first supercomputer, "Galaxy 1", if nothing else, it will come out next year.

The difference between a supercomputer and a mainframe computer, the former is good at numerical calculations, and the latter is good at non-numerical calculations. It's too professional to talk about it, and to put it in layman's terms. Mainframe computers are mainly used for digital processing, for example, major banks have thousands of customers every day to make countless deposits, withdrawals, transfers and other transactions.

It doesn't necessarily require its own computer system, and the computing rate must be very perverted, but the throughput must be very large and very stable. Therefore, the mainframe pursues large-scale computing, not high-speed computing. In order to maintain the stability of the metamorphosis, his accessories will be more conservative, and rarely use the first developed technology.

Supercomputers, on the other hand, are mainly used for scientific research, especially in the defense industry. For example, in the development of nuclear weapons, a large number of numerical calculations must be carried out to simulate the lethal effectiveness of nuclear weapons. When China developed two bombs and one satellite. Using hand-cranked computers and crowd tactics, the used scratch paper can be piled up in several rooms. With a supercomputer, the results can be obtained in just a few days.

There are some things that money can't buy. In fact, in the early eighties. The most advanced computers imported domestically from the West can only be medium-sized computers. And there must be a prerequisite for this – the person in charge of the computer operation must be a foreigner.

For example, in China, a geological simulation operation should be carried out to explore whether a certain plot of land has the basis for oil formation. The Chinese side can only hand over a series of pre-designed variable values to a technician sent by the computer sales company, who will control the computer to get the results and then feed them back to the Chinese side. And the Chinese personnel who paid for the computer were forbidden to approach the computer. It is this kind of humiliation that makes China smash the pot and sell iron to build its own supercomputer.

Supercomputers are a high-investment, low-return industry, and their commercial applications are much smaller than those of mainframes, and few companies that can afford to build supercomputers are actually profitable. Most of the design and manufacturing of supercomputers are supported by national capital.

And the reason why Li Xuan wants to develop a supercomputer. The reasons are similar to those in China. Not long ago, the Oriental Research Institute purchased two of the latest Caryx-MP supercomputers, which were just launched this year, from the American company Cray. It's the fastest supercomputer in the world today, with a rate of 914 million floating-point calculations per second.

The Oriental Research Institute plans to install one of the two computers in the Silicon Valley R&D center, while the other will be installed in the Oriental Real Yàn Building in Hong Kong. These two computers are mainly used to simulate chip computing, so that a large number of design gaps can be quickly found in the chip design stage and improve R&D efficiency.

But supercomputers are a special commodity that is regulated, and the order for the one in the Silicon Valley R&D center was produced by Cray without any restrictions. And the installation and commissioning are completed. The supercomputer of the Hong Kong headquarters was stuck when it was being reviewed by the US Congress.

Li Xuan found two public relations firms to help lobby the relevant lawmakers, and it took a lot of effort to reluctantly get the Senate and House of Representatives review committees to let them go. But Li Xuan's original customized caryx-mp. It was forcibly changed to another model, the Cary-1. The Cary-1 was Cray's last supercomputer, and its fastest rate was only 250 million operations per second, less than one-third of that of the latest Caryx-MP.

Although the price of CARY-1 is only $1 million. It's a lot cheaper than the $1.5 million for Caryx-MP. But including the PR firm's lobbying costs, Mr. Li spent $1.6 million on the CRY-1 supercomputer.

And because Hong Kong is backed by the mainland. The use of this CEY-1 is also subject to the supervision of the Cray Company. It is foreseeable that it will cost its later use and maintenance. It's certainly much higher than that supercomputer in Silicon Valley.

Spent more money, but bought back a product with backward performance. Such a fiery thing naturally made Li Xuan depressed. So he simply decided that Dongfang Electronics would develop a supercomputer on its own.

Of course, Li Xuan, as the boss of a large company, will never set up R&D projects indiscriminately just because of gas, which is definitely a taboo in company management. The reason why Li Xuan decided to let the Oriental Research Institute start the research and development project of supercomputers was not only to fight for a breath, but more importantly, to make a name for the entire Oriental Electronics Company.

It's like the duel of the century between IBM's "Deep Blue" supercomputer and world chess champion Kasparov in the nineties. IBM invested a lot of manpower and material resources in the "Deep Blue" project, but as a result, it only played two six-match battles with Kasparov, with one loss and one win. After that, the "Deep Blue" supercomputer was quickly disassembled, and the unconvinced Kasparov was not given a chance to fight a third time.

It can be said that the "Deep Blue" supercomputer, except for a documentary that recorded the two peak showdowns between artificial intelligence and the human brain, did not bring any benefit to IBM. However, the topic of man-machine warfare has become the hottest topic in the entire 90s, and has been hyped by the global media. The brand promotion value this brings to IBM cannot be achieved by tens of millions or even hundreds of millions of dollars of advertising investment.

After the rapid development of Dongfang Electronics Company in the early stage, it will soon come to the time when it will need to establish its own high-end brand value. Therefore, Li Xuan's requirements for supercomputers are not the same as those of the mainland government, which is working intensively on the development of the first supercomputer. What needs to be solved now in China is the problem of yes and no, and the "Yinhe 1" that will come out next year has an operating rate of only 100 million times per second, far behind the Caryx-MP supercomputer that Cray will come out this year.

And Li Xuan's purpose is to make a name for himself, so an ordinary supercomputer doesn't mean much to him. What he needed was the world's fastest supercomputer to catch the world's attention.

Of course, developing a supercomputer is never easy. First of all, R&D personnel with relevant experience are the basic prerequisites. It is impossible for Li Xuan to mobilize the scientific research strength of the entire country to support it as he does in China. But he also has his advantages, at least without having to cross the river by feeling the stones. The Oriental Research Institute has just found the best candidate who can take on the "Pearl of the Orient" supercomputer project.

"Mr. Chen, I've kept you waiting, there was a car accident in the Hung Hom Tunnel, and it turned out to be very blocked. I'm sorry to be late for the first meeting with you!" Li Xuan said apologetically as soon as he entered the door.

"Mr. Li, hello! It's okay, I also accompanied by Dean Zhao, I just visited the Shi Yàn Building, and the equipment here is very advanced!" Chen Shiqing smiled at Zhao Weiming, who was accompanying him, and shook his head lightly to show that he didn't care.

Li Xuan quickly looked at the man in front of him who was wearing a pair of gold-rimmed glasses and had a very gentle smile. Mr. Chen Shiqing's ancestral home is southern Fujian, and his family went to Taiwan in 1949. After graduating from National Taiwan University, he went to the United States to study and is now a Chinese-American. Most of the Chinese engineers in the high-tech field in the United States are from Taiwan, such as Zhang Zhongmou, the president of Dongfang Semiconductor.

"Mr. Chen, you are welcome to join us, and I welcome you on behalf of all my colleagues at the Institute of Oriental Studies. I believe that in a few years, your choice will burst out with incomparably dazzling light!" Li Xuan said with a smile.

Chen Shiqing's previous role as the chief technology officer of Cray was responsible for the design of the world's fastest Caryx-MP supercomputer. The average person may be relatively unfamiliar with the Cray Company, which was the well-deserved hegemon of the supercomputer industry in the seventies and eighties.

Although IBM, Intel, and Silicon Valley Graphics quickly rose to prominence in the field of supercomputers after the 90s, it is still the most famous supercomputer manufacturer in the United States. Before Li Xuan's crossing, the "Titan" supercomputer manufactured by Cray Company for the U.S. Department of Energy's Oak Ridge National Field Office ranked second in the world in terms of computing speed, although it was surpassed by China's newly developed "Tianhe-2". However, the comprehensive effectiveness of the "Titan" is more powerful than that of the "Tianhe-2", which shows the strong technical strength of Clay.

Although Chen Shiqing is the chief technology officer of Clay Company, the real boss of Cray Company is the founder of the company, Mr. Cray Sr., and Chen Shiqing is just a wage earner. Even if Clay Sr. retires, he may not be able to take over the helm of Clay Co., don't look at the United States' self-proclaimed beacon of democracy, but the ethnic ceiling is everywhere in large corporations. Unless they are founders, very few Chinese Americans can become CEOs of companies.

Li Xuan's condition for poaching the other party was not only to make Chen Shiqing the full person in charge of the "Pearl of the Orient" project, but also to promise that after the successful development of the supercomputer, an independent subsidiary would be established for commercial development. Chen Shiqing will hold a 30% stake in the new company, changing from working for others to being his own boss. (To be continued......)