Chapter 297 The Butterfly Will Eventually Cross the Ocean (6.4k)

"What a bad taste Yuanguang has. Don't you think this scene is very similar to the commercial that Macintosh ran during the Super Bowl?" Wang Ruixin couldn't help but complain as he looked at Chen Yuanguang's head occupying the entire screen.

The ticket was obtained by Li Tianle. NIO and Guangjia Aerospace have a deep cooperation, and many tickets for this press conference were reserved for these cooperative enterprises.

It was easy for Li Tianle to get two tickets for the press conference.

Li Tianle, who was sitting next to him, asked in a low voice: "What advertisement?"

(Classic advertisement in Apple’s history)
"You know '1984' written by George Orwell, right?" said Wang Ruixin.

"know."

"Apple made an ad called 1984, which meant that 1984 would not be the 1984 described by George Orwell, but the 1984 changed by the Macintosh computer.
The ad basically shows a woman in red shorts and a white vest with a Mac logo rushing into a hall full of expressionless, robotic, bald men and smashing a screen showing Big Brother's message with a hammer.

The huge head portrait on the screen is very similar to this scene. If there were large and small TVs around the press conference, and Yuan Guang’s head portrait appeared on the TVs at the same time, it would be even more similar. "

Li Tianle was a little confused and didn't quite understand what the other party wanted to express.

He looked around. Everyone was immersed in Chen Yuanguang's speech. Only the two of them were chatting.

This time, Li Tianle's voice lowered, "What do you want to say?"

Li Tianle knew that his friend was not the kind of person who liked to make random comments, so there must be a meaning behind what he said.

Wang Ruixin said: “If there is a science and technology cult in the country, Yuan Guang is the undisputed leader.

Don’t you think this scene is like a sermon? Everyone is bowing down in worship.

All that's missing is some religious ceremony.

I didn’t realize before that Yuan Guang had this talent.”

Modern history has caused the Chinese to believe in technology. If there really is such a thing as a technological religion, then the Chinese must be the earthly kingdom of this kind of religion.

After listening to this, Li Tianle shook his head helplessly and did not say anything else.

Chen Yuanguang was still saying on the screen:
“Everyone should know that Dongda’s consumer electronics, semiconductors, new energy vehicles and other fields are facing unprecedented difficulties due to the suspension of supply from foreign suppliers.

In the past, we had to rely heavily on imports for chips, especially those with advanced processes. If we relied on the domestic supply chain, we could only produce 7nm chips at best. Topological semimetals, in essence, achieve the equivalent of 1nm by stacking transistors.

The size of its transistor is not really 1nm, but what is equivalent is that in effect, it is equal to a 1nm transistor.

What is frustrating is that the supply of new photoresist, the core raw material of topological semimetal process, has also stopped.

Everyone knows more or less the impact on our industry. This time I just want to say one thing, that is, the impact will not last too long and will be much shorter than Washington expects.

The problem can be solved in half a year at most.

Suppliers led by Xida and Neon, when they choose to withdraw from Dongda Market, the door of Dongda Market will be closed to them forever. "

There was thunderous applause from the audience.

After the applause died down, Wang Ruixin turned his head and looked at Li Tianle: "Look, this is Yuanguang's appeal.

Although Yuanguang did not make it clear what process it used and how it solved the problem of chip limitations.

But the 5,000 spectators present had no reason to believe that Yuan Guang could solve it.

I think Yuan Guang’s nickname should not be God of Light, but rather Leader.”

Li Tianle said as a matter of course: "This is natural.

If we had not been Yuanguang's roommate, we would also worship him just based on his grades.

Even as his roommate, I admire Yuanguang's achievements.

Many things can be said to be the result of luck, such as room-temperature superconductivity, topological semimetals and perovskite batteries. I have seen people on Zhihu or Quora say that these are due to good luck.

I don't believe this excuse. Why is Yuanguang always lucky and not others?

But let's just say this is just luck.

Cracking the laws of virus evolution, inducing the virus to evolve in a specified direction through a broad-spectrum vaccine, and creating a new theory of vaccine design, this can't be luck.

A virus prediction model was established, which successfully predicted the COVID mutation and virus mutation points over a five-year period.

This isn't luck either.

As for the artificial intelligence model for predicting the properties of small molecule metals, many companies I know are still using this model today, and many practical algorithms have been developed based on this model.

Is this achievement a small breakthrough in the field of computational chemistry in recent years? It has promoted the improvement of the status of computational chemistry as a whole.

The latter are enough to prove that Yuan Guang has a solid foundation. It’s just that too many more dazzling achievements have overshadowed these things that prove his basic skills.”

Wang Ruixin immediately said: “I just think this scene is a bit strange.

Maybe it’s because I haven’t been back to China for a long time, so I always want to complain when I see scenes like this.”

Wang Ruixin did not make any further excuses, and Li Tianle also did not say anything.

The video on the stage had finished playing, and after the applause died down, Lin Jia came out to speak:

“I have many identities. I originally thought that my next press conference would be for Space’s new energy vehicle product launch, but I didn’t expect it to be for BiRen Technology’s product launch.

I also browsed various social media platforms, and there are many interpretations about BiRen Technology online. I believe that all the audience present here are also very clear about what kind of company BiRen Technology is.

Just as you guessed, the product we are going to release this time is the graphics card independently developed by BiRen Technology - Longyin I.

We have divided it into two versions: basic version and pro version.

In the ladder running points, the basic version of Longyin I is benchmarked against 3080, while the pro version is benchmarked against 3080ti. "

The main domestic companies that make GPUs are Moore Threads and BiRen Technology. BiRen Technology previously focused on making AI computing cards, which means it did not develop graphics cards for gamers.

Therefore, only Moore's Threads are developed for graphics cards for personal users, and the products of Moore's Threads can be described in one sentence: they are usable.

The most advanced personal product launched by Moore's Law before was only comparable to the 2060. Even because of the driver and adaptability reasons, the running score was comparable to the 2060, but in actual use in many fields it might only reach the level of 1070Ti.

You have to know that 1070ti is a product of 2016.

That was ten years ago.

BiRen Technology has raised this level from 2016 to 2020 at one time.

The applause from the audience became even louder.

You talk about buying an AI computing card, but few people would buy this thing themselves.

But when you mentioned buying a graphics card, everyone was still very interested.

"Do you want to buy it or not?"

"What about you?"

"I won't persuade you to buy it, and I won't buy it even if you persuade me. I want to wait and see."

“I will definitely buy it. Guangshen itself has never produced products for individual users. This should be the first time. I will definitely support it.

I don't think I will be able to afford the car he will produce in the future. Even if I can't afford a car, I can still afford a graphics card.

Besides, Xida is targeting the domestic semiconductor industry, so I will support it with real money.”

When the audience members were discussing individually, most of them just wanted to show their support.

"Longyin I uses a process equivalent to 6nm, but we can actually achieve 3nm," said Lin Jia. "Relying on a domestic supply chain, we can achieve an equivalent of 3nm.

Therefore, next year, we will hold another press conference here to launch the second-generation Longyin graphics card that will compete with NVIDIA's 50 series graphics cards."

It's made nationwide, it's like an aphrodisiac, especially at this juncture.

The main reason is that everyone previously thought that as long as we de-mized it, it would be fine, and this was also a way to solve the problem. But now it seems that even de-mization is not enough, and we have to go all in.

If it is not domestically produced, you will face crisis at any time. After all, it is unrealistic to expect other countries to listen to you instead of Xida.

Even France is a beneficiary of the international order built by the University of the West. They may waver on non-critical issues and may stand on your side for some reasons. On critical issues, they can only maintain a neutral attitude at most.

To a certain extent, the entire industrial chain is domestically produced and is gradually becoming a giant industrial Leviathan, which is also forced.

"The Dragon Song II we just talked about was just a glimpse. Now let's get back to the Dragon Song I graphics card that was released this time. Let's watch two real-life demonstration animations using the Chinese 3D masterpiece Black Myth: Wukong that was released two years ago."

After Lin Jia finished speaking, the screen was divided into left and right sides, both showing the same actual game demonstration animation, and the animation effects were very smooth.

“We can see that without turning on the ray tracing special effects, the basic version of Longyin I can perfectly support the display of Black Myth’s 1080P picture quality, while the pro version can support 4k ultra-high picture quality.

Of course, the premise is that ray tracing is not turned on.

This graphics card does not currently have ray tracing technology."

The press conference lasted about an hour. After listening to the whole thing, Wang Ruixin and Li Tianle went to the nearest Hangzhou cuisine restaurant for dinner.

Wang Ruixin asked: "What do you think is the most important message after listening to the whole speech?"

Li Tianle is clear about this: "With regard to chip technology, the country can achieve 6 nanometers this year and 3 nanometers next year.

To put it bluntly, the time window obtained by Xida this time is only one year, and foreign chip manufacturers only have one year to enjoy the premium.

If they can’t achieve a major technological breakthrough within this year, then the University of Tokyo will catch up.”

“Yeah, that’s definitely something they didn’t think about before making the decision.

Because the domestic lithography machine of Dongda University has only conquered 14nm in public before, and this 14nm is still very watery, it is just de-mization.

From the de-mized 14nm to the 6nm produced nationwide at today's press conference, the gap between them is too big, and the decision-makers in Washington made a fatal strategic misjudgment. "I can only say that Yuanguang is powerful enough to launch a counterattack in just one month." Li Tianle said with emotion.

When he was at Yuanguang University, he was the focus of the School of Chemistry. However, at that time, he became the focus not because of his scientific research achievements, but because of his relationship with Lin Jia. At that time, it was not easy to pick up girls from the School of Economics and Management.

Not to mention that Lin Jia is one of the top five girls in the School of Economics and Management in terms of appearance. If we consider her overall strength, she is the top girl.

Everyone was envious at that time.

Li Tianle lamented that even today, Yuan Guang is still the focal figure among them.

“No, it definitely cannot be completed within a month. It has been prepared a long time ago, but it was only brought out today.

With acquisitions like BiRen Technology and Maolai Optics, personal graphics cards have definitely been ready from research and development to tape-out and then to production.

It was only announced at this point in time, and the changes in industrial and commercial information were only recently exposed.

I always feel that this is a trap dug for Xida in advance. If you don't turn the table over, then we will cooperate well. If you turn the table over, then we won't play together. We can continue to play the semiconductor game without you.

This time we start with graphics cards, which is clearly aimed at Nvidia. If we can conquer graphics cards, then Dongda can come up with alternatives for all types of semiconductor chips.”

"This is a good thing."

I don’t know if it’s a good thing, but it’s definitely a good thing for consumers.

The next day, pre-orders for the Longyin series of graphics cards were available on all platforms, and pre-sales were opened simultaneously on multiple channels including Taobao, JD.com, and Pinduoduo.

The price of the pro version is 1999 yuan, while the basic version is only 1599 yuan.

The graphics card performance of 3080ti is similar to that of 4070, while in the 50 series, it is comparable to 5060.
Compared with NVIDIA graphics cards with similar performance, the price of the Longyin series is about 40% lower.

This price is definitely a great deal for consumers who need to buy or replace their phones recently.

"Do you think it can be sold at this price?" asked an old employee of BiRen Technology.

The senior executives of BiRen Technology have basically been replaced, and the middle-level leaders of the functional departments have also been replaced.

Basically, they were replaced by employees who were airdropped from Guangjia Aerospace.

There is no change in the grassroots staff.

Even those who originally wanted to resign gave up the idea after learning that their own company was acquired by Guangjia Aerospace.

This is equivalent to directly joining Guangjia Aerospace, which is no different from following a mentor to transfer from Harbin Buddhist College to Harvard.

Some bloggers on Zhihu who got offers from BiRen Technology but didn't go in the end regretted it so much, like "I got offers from *** Company and BiRen Technology, which one should I go to?"

This type of question is filled with new answerers asking questions in return: "Did the blogger go to BiRen Technology?"

Everyone wants to see an answer that will make them regretful to satisfy their own bad taste.

"Don't worry, we only need to worry about production capacity, the price is too low.

Even if our pricing is a little higher than Nvidia's similar chips, consumers will still pay for them.

Do you know what domestic premium and patriotic pricing are? Huawei can enjoy them, and so can we. Employees from Guangjia Aerospace are well aware of this.

"Then why is the price only set at 2000 yuan this time?"

"It should be out of caution. After all, this is the first time to make a graphics card."

“I don’t think so. The target audience of graphics cards are young people, who are price-sensitive consumers. They may pay a little premium for patriotism, but if the premium is high, only a few of these consumers will buy it.

Young people who accept Huawei's pricing are a minority, and most are middle-aged men with plenty of money on hand." said another Guangjia Aerospace employee who was parachuted in.

After the official sale started at 10 a.m., the units prepared were sold out in minutes.

Fifteen minutes of revenue is equal to the revenue of BiRen Technology in the past year.

"You're right. I underestimated Yuanguang's appeal. I thought the sales of graphics cards would be very limited, and they were not market-proven graphics cards.

It is surprising that so many consumers place orders blindly. "

"Domestic demand is huge enough, and coupled with everyone's frustrated mood over the past month, it will take only a few minutes to digest this amount of orders."

At about the same time that day, OPPO officially posted a Weibo:
"After three years, the butterfly returns"

The accompanying picture is a photo of the front of a chip.

Many people in the digital world know what this means at a glance.

In the first quarter of 2023, OPPO's Zheku team successfully taped out TSMC's 3nm chip. Everyone thought that China would have a second self-developed chip manufacturer. The first was Huawei's HiSilicon Kirin.

3nm will be a high-end chip in 2023, comparable to the flagship chips of Apple and Qualcomm.

Before domestic consumers could wait for more information to be revealed, what they waited for was the news that Zheku was disbanded, retaining only the core architecture team.

Three years, return, the meaning couldn't be clearer.

"In the past three years, there have been frequent reports of Zheku's reorganization. It was not until the end of 24, when Huawei launched the Ark chip, that Zheku was officially restarted.

However, this Zheku is different from the Zheku in 23. The chips made are different, the talents required are different, and it is impossible to directly use the architecture of the tape-out in 23. Everything has to be started from scratch.

Previously, the team size of Zheku exceeded 3000 people. Before the disbandment, it had launched two self-developed chips, Mariana MariSilicon X and Mariana MariSilicon Y. The former is an NPU chip with a 6nm process, which was very advanced at the time.

From the perspective of technical strength, it has been fully proved that Zheku has no problem at all, and it is very likely to make chips at the same level as Kirin or even better.

Dissolution is common, but directly dissolving a business unit is uncommon, and directly dissolving a company with 3000 employees has never happened.

Generally speaking, if a chip company really can't survive, it can lay off employees little by little.

Foreign chip companies like Shenhai basically lay off 50% this time and another 50% next time. However, Zheku laid off 3000 people at once and the compensation was very sincere. It was not because of poor management at all. To put it bluntly, it was because of external pressure.

Once you do it, you will have to face the government in Washington. Either TSMC will not manufacture for you, or Qualcomm will not sell chips to you.

The restart of Zheku two years ago was also based on self-research. Who would have thought that Washington's butcher knife would still be cut down, and this time not only OPPO would not be spared, but all mobile phone manufacturers would not be spared.

Foreign markets don’t allow you to do business, so you can only do business in the Dongda market.

Originally, these domestic mobile phone manufacturers were very cautious. Although they had established their own chip teams, they were thinking about gradual replacement and first launched their own topological semi-metal chips on their flagship phones.

After all, this thing does have better performance. Its power consumption, heat dissipation and performance are all better than traditional silicon-based chips. Huawei has this thing, and other mobile phone manufacturers cannot compete with Huawei in the domestic market if they do not.

Even so, all they want is for us to continue buying Qualcomm chips so as not to provoke America.

In the end, the butcher's knife was finally cut, and it was even more targeted than they thought. You have to buy Qualcomm chips and give up overseas markets.

It just so happened that at this time, domestic chip manufacturers offered an olive branch. Foundry, right? We can do it, but we can only do 6nm. Do you want to do it?

OPPO already had Zheku and had experience in wafer tape-out. Anyway, the University of Tokyo would not allow us to enter the overseas market, so I might as well cooperate with a company in the field of optical armor.

As for another concern, if OPPO insists on developing its own chips, will Qualcomm stop supplying them completely? I don't think there is anything to worry about. You can't just target OPPO.

If you target all Dongda mobile phone companies, wouldn't that mean you're giving the market to Huawei? In addition, Apple and Samsung can forget about entering the Dongda market.

So I think OPPO was forced into a corner this time and was determined to go against Xida.”

Analysis by an OPPO blogger.

But there is one thing wrong in his analysis, and that is the process.

Frontier Technology's foundry product for OPPO is not 6nm, but 3nm.

The chip design that was taped out in 2023 will be used directly in 2026. It was successfully taped out that year, so nothing needs to be changed this time, only the process needs to be debugged.

OPPO senior vice president and former OnePlus founder Pete Lau posted a long Weibo post, which can be regarded as a semi-official channel of OPPO:

"The chip that OPPO is going to release this time is named Butterfly, Butterfly-9400S, which uses a 3nm process. It is exactly the chip that Zheku failed to produce three years ago.

When developing this chip, our Zheku team worked 7x16 hours a week, carefully studied a large amount of competitor data, and referred to a large number of demand use cases. The entire architecture design team completed a highly competitive architecture design in just 4 months and completed the issuance of the performance feature set within 6 months.

This is not an easy task, as the SoC that never came out had features that are still outstanding even today.

It adopts the latest Arm Cortex-X series CPU cores and the latest GPU cores, rich CPU L1/2 cache and huge L3/SLC cache, DDR latency below 100ns, asynchronous MTE, AI DVFS/FI/SR), powered by a series of efficient internal custom IPs.

On the night when OPPO decided to disband Zheku three years ago, I was in my office in the OPPO office building, playing Faye Wong’s song over and over again. The song was called "Butterfly".

The reason why I played this song is because there is a line in the song: Just like a butterfly cannot fly across the ocean, no one has the heart to blame.

We gave up Zheku and self-developed chips because of the pressure from American. This pressure forced us to choose to destroy our five years of hard work with our own hands.

This kind of pain cannot be described in words. My thought at that time was that perhaps the butterfly will never be able to cross the ocean.

Coincidentally, the only chip that Zheku has developed and successfully mass-produced is called Mariana. The Mariana Trench is the deepest trench on Earth. Our idea when we named it this way was that we wanted to do the most difficult thing in the world and explore the limits like going to the deepest part of the Mariana Trench.

The result is that the butterfly cannot fly across the ocean, Zheku cannot fly across the Mariana Islands, and even if the tape-out is successful, there is no way to launch our self-developed 3nm mobile phone SoC.

Later last year, when we were preparing to develop our own chips again, this time using a completely new process, we faced the same dilemma. I was almost in despair. Do you know how painful it is to encounter such force majeure continuously, which resulted in the inability to launch chips that could obviously be launched?
I once thought that OPPO was naturally in conflict with self-developed SoC. Fortunately, before we disbanded Zheku, Guangjia Aerospace contacted us and asked us to wait. As a result, our own domestic chip industry chain was able to achieve a 3nm process.

Today, the chip that failed to come out that year will be launched within this year, and this time it will return through domestically produced technology.

The butterfly will eventually cross the ocean! "

OPPO has a natural advantage because they have already been successful and can just use that design directly.

This was almost the last straw for OPPO's chip R&D team, so after receiving the invitation, they wasted almost no time and the entire team moved to Mianyang to start work.

After the successful tape-out verification, OPPO, from the official to the executives, began to make public statements.

(End of this chapter)