Chapter 168 Direction
“It’s really stormy.”
The weather in Yanjing is very good, the sun is shining brightly, Xiaomi Motors’s start is far beyond expectations, both in terms of sales and reputation, it is better than them The best expected is even better.
The main reason why we chose to make a coupe instead of an SUV, which is the most popular model in China, is path dependence. Xiaomi mobile phone was originally aimed at Apple’s sinking market. The car is targeting Tesla’s sinking market.
Unlike fifteen years ago when the price of Xiaomi mobile phones was halved and then slashed again compared to Apple, the price of Xiaomi cars is not much lower than that of Tesla. The main focus is to provide a better experience at the same price.
As for whether this can be achieved, the executives were not sure before it was officially launched on the market, and Lei Jun himself was not sure either.
Now Xiaomi Motors has delivered better results than expected, and Xiaomi’s stock price in Hong Kong has soared. However, neither the good weather nor the recent beautiful results of Xiaomi Motors can calm Lei Jun’s heart.
Because at the Huawei press conference yesterday, Ark was like a bolt from the blue, which made Lei Jun feel exhausted.
As a participant and witness in China’s smartphone industry, Lei Jun read a clear signal from yesterday’s press conference: The mobile phone industry is about to start a war again.
Domestic 1nm chips are the clarion call. In the future, the mobile phone market will be divided into two camps, domestic chips and Amerikan chips.
In this environment, it is absolutely impossible for Amerikan chips to gain recognition from consumers in the domestic market.
In recent years, Chinese consumers' enthusiasm for supporting domestic products has far exceeded imagination. As long as domestic chips can achieve 80% of the experience of foreign chips, most consumers will prefer domestic chips.
Not to mention that the performance of domestic chips is so far ahead that there is no way to beat them.
Xiaomi’s Ultra series of mobile phones are completely incompetent against Huawei’s Mate and P series. In the past, they could rely on differentiated competition and the experience advantages brought by Qualcomm chips to win the market. What about the future?
Lei Jun’s judgment is similar to that of OPPO. Next year’s 1nm Ark chip will definitely be decentralized to the Mate and P series. By then, Xiaomi’s high-end mobile phone strategy will completely fail.
Not to mention Xiaomi, even Apple cannot hold on to the current market. Otherwise, why would Apple's stock price plummet by more than 10%.
As for joining the domestic camp and giving up Qualcomm, this decision is even more difficult. Xiaomi and Qualcomm are much more closely tied than other manufacturers. It can be said that Xiaomi was founded by Qualcomm.
Not to mention that there is a share swap agreement between them. When Xiaomi gave up the Pengpai chip, it was because the chip was really difficult, and also because Qualcomm lobbied for it.
Xiaomi is the most difficult decision among domestic manufacturers to make.
This can also be seen from the progress of self-research of chips by various manufacturers. OPPO once had Zheku, VIVO had self-developed ISP chips, not to mention Huawei.
Only Xiaomi's ThePaper S1 fell into a long stagnation after its failure. In the past two years, they have begun to join forces with MediaTek to develop mobile SOCs.
But the chips they are currently developing belong to the mid-range market, which means they will not affect Qualcomm’s core interests.
"Mr. Bin, do you think it was a wrong decision for us to give up the Paper S1?" Lei Jun was very melancholy.
Lin Bin is one of the earliest founders of Xiaomi. He left his position as a senior executive of Google Greater China to start a business with Lei Jun and made great contributions to Xiaomi.
When Xiaomi gave Lei Jun 9.9 billion yuan in stock incentives, it was Lin Bin who stepped forward to answer questions from the outside world, which roughly meant that this was a decision made by several of their directors without telling Lei Jun.
Believe it or not, that's what he said.
Lin Bin said: “No, self-developing SOC is a thankless task. We gave up on ThePaper S1 back then because it was difficult to even make a profit, let alone make a profit.
< br>No one could have imagined that China would be the first to break through 1nm in chip manufacturing today.
The decision we made back then was understandable.” In order to give the market confidence, the plan of Lianxin, a subsidiary of Datang Telecom, was packaged and purchased. It is natural to give up the plan if it is not done in person.
"The chips we are jointly developing with MediaTek will probably be outdated by the time they are built.
We now need to make an urgent U-turn and redesign the chips using topological semi-metallic materials," Lei Jun said.
Lin Bin thought for a while: "In fact, we don't have to develop it ourselves. We can choose to cooperate with some domestic manufacturers and use their SOC."
Lin Bin continued: "We are now The front line is too long, and if resources are invested in the chip, the overall report will not be very good.
And topological semi-metal is a brand-new material, its yield rate, tape-out cost, and chip design logic. These are unknowns and the time we need to invest will be very long
Huawei is already too far ahead in this regard. It would not be wise for us to catch up. I think we should focus our resources on areas where we currently have advantages, such as Xiaomi Motors.
What’s more important is the procurement project memorandum we just recently signed with Qualcomm. We will purchase US$2 billion in chips from Qualcomm in three years.
If topological semi-metal is used for self-development, and the high-end market uses self-developed chips, then these US$2 billion chips will never be used up. ”
Lin Bin obviously doesn’t want to do self-research anymore. In his opinion, it’s great if self-research is successful, but the chance of success is too small.
Once the financial report fails, how can it be related to the capital market? Explain?
So there are advantages to not being listed, which allows you to continue to do the right things in the long term.
"Mr. Lei, I think we should ask Yuan Guang for his opinion. As the core R&D personnel of the technology, he should be the most authoritative about the future development of the technology." Lin Bin finally said.
This statement feels like why not ask Aladdin about the magic lamp.
After hearing this, Lei Jun nodded repeatedly, "I have wanted to call and ask for a long time, but I never thought about what to ask before. I will call when I think about it.
< br>
But his answer may not be what you expect. "
After hearing this, Lin Bin sighed with emotion: "From past experience, it is best to bet on both ends.
It is impossible for the country to let Huawei dominate in mobile phone SOC. Not to mention that the country will not agree, none of the domestic smartphone manufacturers will use Huawei's chips.
Therefore, there will definitely be several chip design companies similar to HiSilicon Kirin in China, such as Unisoc, which has a certain market share before. We can wait for their solutions to mature and directly purchase their chips.
Qualcomm is not offended either.
It’s just that the world is changing too fast now, and I’m not sure whether I should invest more in chip research and development, so I think it’s better to listen to Yuan Guang’s opinion, and it’s never wrong to listen to God’s opinion.
If he suggested that we do our own research, that would be a reassurance. "Lin Bin finally teased.
After dinner at about 7:30 in the evening, Chen Yuanguang received a call from Lei Jun. He smiled when he saw the caller ID. It was a little later than he expected. .
“Brother, what’s the matter? "
Lei Jun's distressed voice came from the other end of the phone: "Junior brother, you have given us a problem this time.
This question is too difficult and we can't solve it, so I'm here to ask you for help. "
Chen Yuanguang didn't mince words either: "Is it a topological semi-metal? "
Lei Jun: "Yes, I was terrified when I watched the Huawei press conference. I felt that we would be killed by Huawei soon.
Huawei itself is stronger than us in some aspects. Previously, their shortcomings were in chips. Due to Americen's restrictions, their chips could only be the same as Qualcomm's two years ago.
After the chip achieves breakthrough, our competition with Huawei will become very difficult. What is the concept of 1nm mobile phone SOC? Five to ten years ahead of us, that’s an exaggeration.
I couldn't sleep all night last night because I was so stressed.
I would like to ask how quickly it can replace silicon and how it performs in terms of cost? ”
This is a problem that goes straight to the core. If the cost of topological semi-metal is lower than that of silicon, then silicon-based chips will completely lose the market even in the low-end and mid-range.
After listening, Chen Yuanguang's mind went back to three months ago. At that time, they were just happy to find the suitable design logic for the light source system. There were still many problems to overcome in the face of mass production.
He remembered clearly. One day after he had a meeting with the Guangjia Aerospace space station design team, Liang Mengsong came to him and said he had a brand new idea:
"Yuanguang, why do we have to apply silicon-based chip technology to topological semi-metals?"
After listening, Chen Yuanguang asked doubtfully: "What do you mean?"
Liang Mengsong was excited Dao: "As the FET process gradually approaches its limit, the industry has conducted a lot of exploration, trying to find ways to break through the process limit.
This includes new materials. Replacing silicon with topological semi-metal as the chip raw material is just One of the many methods, and another material that is considered to have the potential to become the basic raw material for next-generation chips, is 2D material.
Like topological semi-metal, it cannot directly use silicon-based chip technology. Unlike topological semimetals, which remain entirely in the laboratory, the industry has made a lot of exploration into the physical properties of two-dimensional materials
It has moved beyond the laboratory stage to wafer-level exploration.
One of the most mainstream methods for two-dimensional materials at present, the CVD method is very suitable for us to use in the secondary processing of topological semi-metals.
It was originally a method developed to solve the sensitivity of two-dimensional materials and solve the yield and stability of two-dimensional materials.
Use orientation control to select a suitable substrate that matches the crystal lattice, and then grow a film with high crystallinity, complete lattice structure, and controllable lattice orientation.
In order to adapt to topological semimetals, we can use metal-compounds as precursors to achieve a uniform supply on this substrate, and adjust the final suitable substrate through the combination between the film and the substrate. ”
After listening to it, Chen Yuanguang realized that this was a solution. The important reason why everyone had been slow to advance before was that there was no direction.
The synthesis route in future time and space is not applicable to the present, such as using a solid-state atomic spin manipulation platform to build a topological semi-metal chip and realize quantum sensing on the chip.
Where to find solid-state atomic self-selected operating platforms in the first half of the 21st century.
Even if this is already a backward technology four hundred years later, chip production has developed into an out-and-out black box for most humans, almost completely relying on the self-operation of AI.
This backward technology is also out of reach at the beginning of the 21st century.
After Liang Mengsong had a keen insight into the application points of current cutting-edge technologies, their advancement speed around topological semimetals can be described as a thousand miles a day.
With direction, specific technical difficulties can easily find theoretical foundations in the future time and space.
After they used the existing process to create the first topological semi-metal chip, he and Liang Mengsong both knew that the world would change.
(End of this chapter)