Those Years When I Was Forced to Top the Forbes List

Chapter 369 Silicon-based Breakthrough
Ten minutes after Masayoshi Son left.

In the conference room at ARM headquarters, Lin Bin firmly stated, "Your price is too high. If you insist on $1.6 million, based on ARM China's current revenue, it will take us almost 30 years to break even. Under such conditions, we should rightfully have control over the company."

The deal has been reduced from $200 million to $160 million, and Gu An has also "conceded" on many terms.

Warren on the other end firmly shook his head and said, "Your calculation method is flawed. The permanent architecture licenses and modification rights for ARMv6 and ARMv7 are worth $15 billion in themselves. If we cannot take the lead, negotiations cannot proceed."

“But you only gave us the right to sell in Greater China. This restriction has already diminished the value of the ARM architecture. The value under this clause is actually just a virtually blank market. We need to develop a blank market ourselves. For this, ARM China has to pay 50 million US dollars in R&D every year!”

Goto Miki said indignantly.

Warren shrugged and smiled confidently, "But you've guaranteed the absolute security and technological advantage of your system's architecture for the next five years, haven't you? You'll also learn a lot of technology to optimize your 'KOS.' We'll provide you with support in that regard."

ARM, on the other hand, became more persuasive and persuasive.

The apparent weakness revealed by China seems to stem from its relentless efforts to ensure "system security."

ARM's management did the math.

The depreciation cycle for ARMv6 and v7 architectures is only 5-6 years.

The Chinese side will have to make up the $50 million annual R&D expense if the joint venture does not generate sufficient profits.

The research and development results generated by this funding are shared by both parties.

If the research on technologies such as physical IP library optimization and 45nm PDK process design kits were conducted in China, the final results would be stored in the standard cell library at ARM headquarters. The Chinese side would only have access to IP integration and verification, and the patent usage rights they obtained would be restricted by other ARM patents.

As long as they maintain control, they can dictate how R&D funds are used, making it an absolutely worthwhile business.

As for the so-called "Greater China" market, they believe there is indeed not much room for sales. It's not that China lacks a consumer market, but rather that almost all of China's chips are designed by the United States and Europe. As long as ARM China's sales scope is controlled, the actual sales will be generated at ARM headquarters.

Can China really produce any outstanding design results that can be monetized within a few years?

impossible.

This is a one-off deal.

The Chinese side provides a one-time protection fee to purchase membership, and then pays an annual membership renewal fee of $5000 million. After ARMv7.1 is upgraded again, they will decide whether to continue licensing subsequent versions based on the Chinese side's sincerity in cooperation.

In their eyes, this Chinese company is actually very clever. By using a joint venture, it has ensured that their systems will not be subject to demands from powerful manufacturers to cut off ARM's supply due to competition with European and American rivals. This is a very wise risk-avoidance behavior.

But for ARM, this will be another successful example of exporting technology to developing countries, taking full advantage of their strong technological needs.

At that moment, Lin Bin's phone vibrated.

He glanced at the incoming call, immediately apologized, got up, and walked out the door.

He glanced around and came to the area behind a surveillance camera, intentionally "avoiding" the camera's view.

"Hey."

A calm voice came from the other end: "It's time to wrap up. Let's sign the contract once the price is agreed upon."

Upon hearing this, Lin Bin immediately raised his voice.

"I know."

"I told you, you were too impulsive, you shouldn't have come!"

"51% is impossible! How could they agree to that? I can only try my best to lower the share price!"

“Annual R&D expenses are also necessary! If the Chinese joint venture doesn’t have some technological accumulation, we won’t even have any bargaining power for future ARMv8 and v9! They can just abandon this joint venture if they want! If we have some technology in hand, at least we can use the compulsory licensing provisions of Article 5 of the Paris Convention to obtain the technical documents generated during the bankruptcy liquidation of the joint venture and continue to negotiate with them, right?”

"Hey, being so forceful when negotiating with foreigners won't work! Just listen to me!"

"Alright, alright! Give me one more day! If I still can't reach an agreement, then it's your turn!"

The person on the other end of the phone remained silent throughout; it was only Lin Bin performing.

In ARM's monitoring room, an employee who understands a little Chinese puts headphones on his face to listen carefully and translates in real time to the person next to him. If there is anything he doesn't quite understand, he can go back and listen again.

The standing supervisor smiled.

This stupid Chinese guy thinks that just because he's hiding from the cameras, we won't know what he's doing?

Our surveillance system can pick up sound!

I heard it very clearly!

A few minutes after Lin Bin returned to the negotiation room, Warren's phone also lit up, displaying two long text messages.

Warren observed silently, his mind racing.

Sure enough, there were internal conflicts among them.

That cunning Chinese man has been trying to take over the negotiations and cancel the terms he previously agreed to!

Within minutes, a text message arrived from founder Robin Saxby.

"That Chinese guy threatened me through Guyon of PwC, saying he wanted to ban the ARM architecture in China. He was a bit furious and kept trying to contact me directly. It seems you did a good job, but you need to hurry. I haven't been to the company for several days. I have some other work to do."

Warren smiled slightly, pleased with his negotiation results. Looking at Lin Bin, who had a worried expression, he felt that the other party was quite kind-hearted.

"To be honest, Mr. Lin, you won't get control of the joint venture. If you give up your control and fully accept the terms of access to the technology, I can consider the price of the equity."

He spoke casually, clearly enjoying the feeling of being in control of the negotiations.

Lin Bin, sitting opposite him, remained silent for a while, seemingly holding back his anger, before standing up and speaking loudly: "Fine, I can concede on the dominant position! But I cannot accept complete access rights. In the four areas of IP core development, multi-core bus design, power management algorithms, and security architecture, we must raise the level of access! Furthermore, the intellectual property rights of the technologies developed must be held by the joint venture! You can share technical documents with headquarters for the purpose of developing other technologies, but we must also have the same rights! This is a fundamental right that we should receive based on our investment! Otherwise, this would be an international hegemonic clause! Not only will we be angry, but the Chinese people will be angry too!"

Even from several meters away, Warren felt like the other person's saliva was about to spray onto his face.

But at least they finally budged on the crucial issue of control, which must have made the other side anxious.

“Mr. Lin, we are not maliciously appropriating your intellectual property rights. We will respect the intellectual property rights you should possess. I was just saying that you might not be able to use the condition of ‘licensing technology to a third party,’ since the technology you obtain will not be complete and will be deeply tied to our other proprietary technologies, which may not necessarily have economic value. Moreover, you also want the joint venture to continue to obtain our next-generation licenses, right? You must show sincerity and guarantee the security of our overall technology.”

“I know! But I must guarantee that you won’t discard our cooperation agreement like a piece of waste paper a few years from now! If you do, even if we don’t use these technologies, we’ll sell them to other third parties who might need them!” Lin Bin argued confidently. Warren already understood Lin Bin’s intentions from the phone call; he was simply seeking reassurance. These technologies were tied to their other technologies, so selling them wouldn’t have much value; it would only allow other companies to glean insights into their technological thinking.

ARM updates and iterates very quickly; in fact, after a few years, it may not even be worth spying on anymore.

“Alright.” Warren flipped to the pages of the draft contract in front of him that contained the technology control clauses, and his tone softened. “We’ll discuss the access level further.”

Lin Bin was still a little nervous: "I hope we can work overtime tonight. My board of directors doesn't give me much time."

Warren immediately nodded and gave him an approving look.

“Mr. Lin, we can make an exception and work overtime for you. Your board of directors is very wise to choose people who know how to work to negotiate with us. You don’t need to worry about your Chen having a conflict with you. After the negotiations are over, I will send an email to your board of directors on behalf of the ARM board to express my appreciation for you. Your board of directors and shareholders will definitely know your work ability.”

Since Gu'an has no listed companies and its board of directors is not transparent, coupled with Lin Bin's intentional misleading, Warren naturally assumed that Qidian Technology also operates like an American company.

The group of people on the opposite side, however, were trying very hard to suppress their laughter.

Did Chen and Lin Bin have a conflict?

If that's the case, what's the point of sending an email to our board of directors?

The board of directors didn't dare to say a word!
The cooperation agreement was successfully signed early the next morning.

$1.3 million, which is exactly 10 billion RMB.

Lu Xiaochun went to arrange the payment, while Lin Bin and Goto Miki chatted about yesterday's negotiations in Chen Xuebing's hotel room, laughing heartily.

After laughing, Chen Xuebing said with a straight face, "I have to say something to you guys! What do you mean I am the board of directors? I can only represent the board of directors! If the ARM board of directors really sends an email to praise you, as the chairman of Qidian Technology, I will still have President Lu personally commend you by the board of directors! This is also a work performance, and it should be included in the management evaluation!"

Lin Bin smiled, not caring.

“ARM didn’t show much pressure in this joint venture, which means they can handle the British government. I think the key point is that they have the initiative. But if you can handle the conversion between 49% and 51%, how exactly do you plan to do it? Under these terms, if we get 51%, we will likely face objections from the British government.”

Chen Xuebing waved his hand: "You don't need to worry about that. We're going to the United States tomorrow. You need to focus on two things. First, fully launch the work of the US data center. Second, recruit chip design talents and find a way to cultivate a strong technical team. We don't need to send a large number of talents to China, but we need to be able to provide technical support to China and conduct remote teaching so that the outstanding design talents we recruit can have long-term access to the world's most advanced chip information and technology."

Lin Bin slowly shook his head and said, "I know you are anxious. Our chip design personnel shortage is indeed too great. However, more than 70% of the core knowledge of chip design still relies on process iteration feedback. Purely theoretical teaching that is detached from production line practice is not very effective. In the end, we still need to involve the domestic IC department in practice and bring questions to class so that the US can provide effective guidance."

"Moreover," he added, frowning, "the United States has a data flow ban, which we'd better not violate, otherwise we'll easily be targeted. In other words, it's almost impossible for us to legally involve foreign personnel in core domestic design. Also, top American design talents earn over $250,000 a year, and each person is also equipped with over $100,000 worth of work equipment. If there are no projects for them to participate in, the cost-benefit ratio is close to zero."

Upon hearing this, Chen Xuebing scratched his head, realizing that training chip design professionals was indeed quite challenging.

Singularity's foundation was too weak. In 2006, the statistics of chip designers in mainland China were only 20,000, and there were probably less than 500 people who had participated in the design of 65nm chips. Most of them were in foreign companies such as Google. There were probably only 200 to 300 people who were truly domestic, and Spreadtrum accounted for half of them.

This is because Spreadtrum has the profitability of its turn-key business to support it, and it also acquired the ADI team, taking the lead in trying to design a 65nm DPS coprocessor. Moreover, this was achieved by sending personnel to the European branch. If there is any output, it still needs to be contracted to manufacture overseas and then exported to China for sale.

In China, it is impossible to obtain authorization for 65nm design tools. High-performance computing servers for EDA and simulation verification are not exported to China. Even TSMC and UMC's 65nm PDK (process design kit) are prohibited from being provided to mainland companies. SMIC is currently developing its own 65nm PDK.

Spreadtrum has done relatively well, but its own Singularity is really struggling. When it first established its Shenzhen office, it recruited a group of IC designers, but so far it has not produced anything concrete. It has only accumulated some experience through the design of 130nm and 90nm, and has a talent base of more than 200 people.

Compared to domestic companies, this is definitely an advanced team. But without comparison, there is no harm. Compared to Spreadtrum's smooth development, Singularity seems to lack the genes. Despite investing a lot, it is struggling to make progress.

Chen Xuebing wondered if Spreadtrum's idea of ​​using US data centers to train talent stemmed from its US technological foundation, which he saw as a desperate measure.

“I don’t care about any of that. Just recruit people. If there are no projects, create two for them. If remote teaching doesn’t work, send people to the United States to participate in these projects. It’s our own company. At least we can make our own decisions about how to cultivate talent through projects.”

Lin Bin laughed when he heard this.

"Would you like to hear my opinion?"

Chen Xuebing glared at him: "Why won't you listen to what I'm saying?"

"Okay, actually I've made a plan, and it's with President Lu." Lin Bin said slowly:
“What you said about sending people to the United States to participate in advanced design learning is indeed feasible, and an offshore team is a must. But what you need is to cultivate a large number of talents, and you need a safe and stable environment, so only the top talents can be selected to go to the United States, and most of the training steps still have to be carried out in China.”

"The domestic training approach is to rapidly commercialize the 65nm process."

Chen Xuebing was a bit dumbfounded: "65nm, you can't even participate in the design, and you want to commercialize it? Are you dreaming?"

"I wasn't dreaming," Lin Bin chuckled, giving his answer: "Reverse engineering, physical disassembly."

"Domestic software capabilities are insufficient for reverse engineering 65nm chips, but hardware capabilities are possible. We use nitric acid and concentrated sulfuric acid to open the encapsulation, remove the passivation layer, then polish it layer by layer, perform layer-by-layer imaging, and reconstruct the chip's structural layers and restore the circuitry."

“We purchase some mobile phones equipped with 65nm chips through Hong Kong channels for this kind of physical research. Texas Instruments’ OMAP chip is very suitable. They are leading in ARM architecture research, and several 65nm products will be launched soon.”

"By analyzing the physical structure of the chip, then using the ARM reference design for independent modification and module innovation, and then verifying it through tape-out, we can almost obtain the entire process of accumulating design experience, and we can also follow the correct experience of others to avoid repeated trial and error."

"We will first cultivate basic talent through 130nm and 90nm design. The second stage is to physically break down 65nm. Teams with ideas will go overseas to design, tape out, and produce. Good products can be contracted out from overseas branches and sold back to China, while bad ones will be shelved. The entire IC team is divided into three groups, dozens of groups, to carry out the work in an orderly manner, forming our own talent training pipeline. This way, we can complete the training of the design team, and resources can be reused. Talents can also be invested in batches to advance. In three years at most, we will have an orderly team of at least a thousand people."

Chen Xuebing remained silent for several seconds after hearing this.

Why didn't you say so sooner??

"Didn't you say you wouldn't let me participate in the management of the IC department?" Lin Bin asked with a light laugh.

"."

Lin Bin laughed again: "I told you, I've already given the plan to President Lu. Besides, it's useless to bring it up too early. We didn't have such a large-scale mobile phone layout before, and we couldn't sell our technological output. Let me do the math for you. Physical reverse engineering requires buying a lot of advanced imported equipment, which will cost at least tens of millions of US dollars. The overseas design department needs to gather at least 200 international senior engineers to drive the design process, which will cost another 50 million US dollars a year. Tape-out verification subsidies will cost at least 30 million US dollars a year, right? We also need to develop our own EDA tools to solve the fundamental problem. My idea is to acquire some non-sensitive American EDA startups, plus some strategic patent acquisitions to advance this work. If we do well, within five years, we have the opportunity to catch up with the self-sufficiency and control of EDA within 65nm."

"If you really want to do this, you'll need to invest over $300 million annually in chip design alone. And we'd better also invest in a domestic wafer fab to get priority access to advanced processes. Even if I told you these ideas earlier, you wouldn't have the money, and you don't even have 65nm reverse-engineered products, so you wouldn't know where to start. Different stages of company development require different approaches. Now that we've reached a higher level of strength, we can change our approach to talent development."

Chen Xuebing felt a surge of emotion.

Investing in wafer fabs was already part of his plan.

He simply never imagined that chip design could be trained in this way.

He couldn't help but move closer to Lin Bin, patted his arm and said, "Hey, if you're going to talk, just say it straight. If our company has annual profits of tens of billions, how can we take the chip industry to the next level?"

Lin Bin chuckled: "Don't be silly, I knew you were too ambitious. Let's be more realistic and talk about it when you have several billion dollars. I won't discuss this with you now."

Chen Xuebing's face fell.

"Damn it, you think you're so great! Wait till we get to America! I'll make billions and show you!"

(End of this chapter)