Reinventing the Millennium

The ice core is almost depleted of photoresist, which had been expected since November.

In the past nearly half a year, all parties have made great efforts and mediation on this matter, and some progress has been made. However, it can only meet the needs of mature ice core technology and some advanced technologies to a limited extent, and in the end Advanced 16nm is always unavailable for purchase.

First, Japanese manufacturers cannot sell them. Bingxin, a mainland wafer manufacturer with a 16nm process, is also clearly restricted by BIS and cannot buy it even if it is wearing a vest.

Second, Japanese manufacturers do not want to sell. Ice Core has tried to purchase through overseas vests, but there are only a few companies with 16nm technology in the world. JSR Co., Ltd. and Tokyo Onka do not want to take risks at this juncture.

Due to various circumstances, the expected state-of-the-art production line for ice cores has stopped operating.

It’s hard to describe my feelings. Even though Bingxin had prepared for the worst from the beginning, and even though it had repeatedly verified the situation and vaccinated its employees over the past six months, when it actually arrived, many people were still there with it. Last minute.

Bingxin's work tasks during this period are very intensive, and there is no holiday even during the Chinese New Year. Now that it is the holiday time for many people, my heart is not happy at all, but full of heavy.

Fab1-6 is the most advanced wafer manufacturing plant in Ice Core and the most advanced in the world. It was completed in 2013. In 2014, it successfully launched the first generation 16nm equipped with FinFET. It completed the upgrade the following year and continued to launch the second generation. Second generation 16nm process.

Ice Core has achieved fantastic success in the past two years or so, surpassing Taiwan Records and Intel, making the world aware of such a wafer manufacturer from mainland China.

From the establishment of the project at the end of 2003, to the start of construction in 2004, to the honor of winning the first place in the world, the journey of ice core has not been easy. When this hard-won success was destroyed, it made the participants extremely surprised. .

This is true for all Ice Core employees.

Because of this, Liang Mengsong, who led the team to secretly develop the 10nm process at SMIC, made a special trip back to Luzhou to welcome this moment with the shocked employees of Fab1-6.

The production line was stagnant, and calibration, R\u0026D tests, optimization tests and other things that could be done were arranged later, but these were just leftovers.

Liang Mengsong has not been seen in Luzhou during this period. Today, when he was standing in the factory area, he noticed that everyone was gradually looking at him in unison, and he also saw that some people had tears in their eyes.

He decided to say something anyway.

What can I say?

Liang Mengsong was silent for a while, looked around at his colleagues who were also silent, and said slowly: "It's very bad now."

The shutdown of the production line is not just bad, it is a serious event, and the ice core faces severe external measures.

"But it won't be any worse than when it started."

Today's ice cores are on the market, have many manufacturing processes, advanced R\u0026D and an excellent team, have many stable customers, and have the support of the entire industry in the mainland.

In 2003, no one believed in the founding team of Yeluzi. In 2005, even the risky trial production of 130nm by Ice Core had to be completed with the help of SMIC.

The ice cores at that time have made it to the present, let alone the current ice cores. How can we not face the future?

Bingxin's co-CEO was concise and concise, but everyone in the factory could understand his meaning.

Liang Mengsong said the third sentence: "What's more, there is still me now."

He didn't speak impassionedly, and he wasn't good at that. He just spoke plainly.

However, the eyes of those present seemed to light up again.

Liang Mengsong originally planned to end his speech, but after thinking about it, he decided to add a fourth sentence: "...and Mr. Fang."

Mr. Fang is also very powerful.

The meaning was expressed well, but this intermittent method suddenly relaxed the atmosphere. Well, as the most professional non-professional semiconductor person, Mr. Fang is certainly very powerful!

Liang Mengsong walked a few steps, reached out and gave a high-five to the employee closest to him: "Have a good rest, the work ahead will be very hard."

With his actions, high fives seemed to become a ritual, and the employees walked up to Liang Mengsong one by one to complete the final handover before leaving.

When the ceremony was over, the assistant next to him couldn't help but asked: "Dr. Liang, when will the production line be restarted?"

Liang Mengsong shook his head and did not answer the question. He just said, "Let's go back to Shencheng."

He doesn't know when it will be restarted here, but the matter is over and he needs to go back and continue the work at hand.

…

The suspension of Bingxin's 16nm production line is a very serious signal, but because Yike controls the shipment of mobile phone products accumulated during this period, this will not be noticed by the market immediately.

However, this is only a matter of time. In addition to the physical properties of the photoresist, many media are also watching, and websites such as "KILL YIKE" continue to expose the bad news that the Yike Department is facing.

It was already expected, it was already the lowest point.

Now the market is more interested in what else Yike can do. If it cannot respond to this issue, Yike's stock price will obviously continue to fall, and even the negative decline in the shock cannot be maintained.

The Yike system is still very powerful, but no matter whether it is Yigou, Yixin or Douyin, they are all integrated and function in one system, and each is still an independent company.

Whether Yike's stock price or business, it all depends on its own performance, and the growth of the system can only be the icing on the cake.

So in June, when the American media generally claimed that Yike collapsed, it seemed to have played a cutting-edge concept card in an attempt to make a difference in the stock price.

Yike is doing deep learning and has gathered a group of people to do artificial intelligence. This is something that the outside world has known for a long time. Fang Zhuo also posted the performance of the robots in the office on Douyin this year.

However, on June 10th, "China Securities News" suddenly exposed a piece of news. Yike had made considerable progress in the field of artificial intelligence, and even revealed that it had registered a "Taibai" on the Yicheng Go website last month. ” account, he continuously played chess with Go masters from all over the world and achieved 17 consecutive victories.

In these 17 games, there are Chinese world champions Gu Li and Ke Jie, Korean world champions Park Ting-hwan and Jiang Dongyun, and Japan's number one Yuta Iyama, which is full of gold.

Artificial intelligence attempts to defeat humans in chess. This is something it has been doing since its inception. For example, IBM launched "Deep Blue" in 1997. It was the first time that AI defeated the world chess champion in a formal game.

However, Go has always been considered an area that is difficult for AI to break due to its complexity.

Therefore, when the "China Securities Journal" shook off the "Taibai" vest that caused a small stir in the Go circle in May, the news immediately penetrated the niche circles and quickly spread in the technology and business circles.

This time, Yi Ke generously confirmed the authenticity of the news and admitted that such a thing did happen. The reason for the incident was that Mr. Fang was curious about the extent of his research.

Coincidentally, Google launched an AI called "Alfa" in the first half of the year, and successively played and won against European Go champion Fan Hui and world Go champion Lee Sedol, so the use of vests for testing came up.

"How far has artificial intelligence reached?" The reporter asked Yike's project leader Andrew Ng who responded this time.

"Artificial intelligence still has a long way to go." Andrew Ng hesitated for a moment and then said, "But in the field of Go, it should be already above humans."

Such a sentence triggered an explosive effect in hot spots.

Yike is very strong and has great R\u0026D, but online chess games are not formal and the mentality of chess players is also different. To make such a conclusion based solely on the test results, does it look down on human intelligence and the art of Go?

Many professional chess players have spoken out and are not convinced by Ng's remarks.

Chess master Nie Weiping also went to the media to formally comment on the chess player "Taibai" from Yike Company: "I watched 17 games of chess played by Taibai. It is indeed very powerful, but to say it surpasses humans is too exaggerated. There were many low-level mistakes in these 17 games, and there were also moments when Taibai narrowly won.”

"If you really want to say that Taibai can surpass humans, then there must be a formal competition. Otherwise, it is just big talk and empty talk!"

Nie Weiping made a bold move and invited Taibai to fight.

Yike Company happily agreed to the invitation. By the way, it also explained the origin of "Taibai". The company named the AI ​​project "Venus" in English and "Taibai" in Chinese.

Venus is the planet Venus, which is opposite to the "Mars" in the mobile phone Mars. There are also characters like Venus in Western mythology, and Venus in ancient China has "Taibai, the essence of Western gold, the son of the White Emperor, the Lord, and the general." The narration of "an image" is where the name comes from.

Yike not only agreed to the invitation, but also expanded the scope and was willing to compete offline with five chess players, including Nie Weiping, and also set up a US$5 million winning bonus.

For a time, not only was public opinion intense, but chess players from South Korea and Japan who were originally hesitant also quickly signed up, demanding to fight for the honor of mankind.

Soon, the time was set for June 26, when Yike Taibai played against the five top chess players from China, Japan and South Korea.

However, before this day came, there was a little tidbit between the human camp and the AI ​​camp. The former asked for three battles, that is, everyone would win two out of three games, while the latter was only willing to go 1V5.

That is to say, five chess players are in a room each, and Taibai will play against them at the same time. As long as humans win three of the games, the prize money will be equally divided among humans.

According to reports, Nie Weiping asked someone to get Fang Zhuo's phone number specifically for this matter, and called him directly to express his dissatisfaction, thinking that such a move was a humiliation.

They are all top chess players, but Taibai "plays in many aspects" is really too arrogant!

Of course, this is not because Taibai is arrogant, but because Yike is too arrogant!

"Teacher Nie, three battles are too time-consuming. It takes at least three days for one person. Yike has many tasks now and everyone is in a hurry. I asked the company and they said that this will have no impact on 'Taibai'." Fang Zhuo politely responded, "Let's all come together. You're welcome. We are not a Go company, we just verify verification technology."

In addition, he also modified the conditions: "Teacher Nie, as long as one person can win a game, it proves that 'Taibai' has not surpassed human chess players, then the company will issue bonuses, and I personally will give an additional bonus to the winner." How about giving 5 million US dollars to chess players and another 5 million US dollars to the domestic Go industry?”

Nie Weiping finally agreed because Mr. Fang was rich.

However, with the establishment of such final conditions, the strong confidence shown by Yi Ke has also made the human Go camp feel like it is facing a formidable enemy. At the same time, CCTV, the Internet, Douyin and even foreign media will conduct live broadcasts to determine the final winner or loser.

Before June 26, Yike's stock price rose because of this out-of-circle movement, which made short sellers criticize Yike's actions.

Russell of RC Hedge Fund has become the vanguard of shorting Yike's stock. He criticized Yike bluntly: "This is Mr. Fang hyping the concept! Even if Yike wins, what can it do? Can it launch any products by playing Go? This will not improve Yike's worsening revenue!"

Russell is actually right, but Yike's stock price is rising... Although not much, it has stopped the downward trend and has a small rebound.

What can artificial intelligence do?

Before the Go game day, Yike's vice president of technology Elder also appeared to talk about the direction in this area, believing that it will make great strides in the future in search, medical care, education and other fields.

As for what this game can prove...

Elder's answer is: "At least it can prove that we are on a good path. In fact, compared with playing against humans, we are more interested in 'Taibai' and Google's 'Alpha'. Which one is stronger? This is really worth a try."

I was inspired by seeing Google's actions in this regard. Although the boss does have the intention of hype, the strength of technology is also of interest to Yike.

Google then responded that it is willing to learn from each other after "Taibai" defeats humans.

The subtext is that if it doesn't win, there is no need to talk.

The media opinion is very lively, the Go camp is united against the enemy, and the financial market responds. Against this background, on June 26, Yike's "Taibai" played against five human players, Nie Weiping, Ke Jie, Lee Sedol, Shin Jin-seo, and Iyama Yuta, in a live broadcast on all channels.

1V5, multi-faceted, human honor, and a super bonus of 15 million US dollars.

Nie Weiping was defeated first, Iyama Yuta lost in the middle game, Lee Sedol resigned, and when Shin Jin-seo and Ke Jie fell into a long thought during the live broadcast, although many viewers did not understand Go, they seemed to feel the grandeur.

The two top players did not move for a long time, but the live broadcast camera switched accordingly, capturing the expressions of many professional players who were thinking, frustrated or painful, and also captured Fang Zhuo, the head of Yike, who arrived at an unknown time.

Fang Zhuo noticed the camera and gave a polite smile.

Five minutes later, Shin Jin-seo stared at the chessboard blankly and resigned.

Ten minutes later, Ke Jie had tears in his eyes, announcing that the last game of the human camp was completely lost.

The Go project full of human wisdom ended with a seemingly calm game process amid heated public opinion. The final result proved that Yike had full confidence to take $15 million...

The Go camp was in turmoil, and public opinion once again enthusiastically discussed the progress of AI. Yike's stock was stimulated by this heat and rose again.

Not only that, Yike also launched a game acquired the next day, named "Human Fall Flat" to suit the occasion.

The Wall Street Journal reported the relevant events with great interest, publishing Fang Zhuo's devilish smile, the turmoil in the Go camp, the appropriate game, and Yike's stock price rebound.

Its final evaluation of this was: "Mr. Fang rarely showed his hype of concepts. He is obviously very good at it, but he didn't use it often in the past. However, even Mr. Fang started to hype it, and Yike's real situation may be worse."

Such an evaluation did not dampen the rebound of the stock price. The popularity of AI was hyped up, and Yike was obviously praised. Its investment in this track and team building were also dug out, proving that it was not a temporary idea.

Fang Zhuo's main purpose for this matter was to make the investment heard. After all, he wanted to see what level the company's research and development was. Of course, he also hyped up the stock price by the way, so as not to let the short sellers leave the market in advance.