Reborn in 2010, the AI ​​in my mind is too powerful.

Chapter 51 Buying a house and recruiting people at the same time

In the days following her return from JPMorgan Chase, Chen Yiran began a period of intensive video calls with her parents.

The family began to carefully consider the housing issue; deciding where to buy was a major decision. Chen Yiran carefully considered the financial aspect and finally persuaded her parents to buy two properties, so that she could live there if she went back to live there.

This raises the question of whether or not to stay in the United States in the future, a decision that the parents are still struggling with.

At this time, America was far from showing signs of fatigue, and even when it did eventually show signs of fatigue, it would be more welcoming to the wealthy.

But after all, he's worked for a state-owned enterprise his whole life, and his parents' education was paid for by the government, which brings us back to the same issue. If Chen Yiran can truly change something, returning to China seems like the only way to truly remember his roots.

The final conclusion is to buy two houses first, and also buy one in the US; we'll see what happens in the future.

Chen Yiran had previously told the beautiful Emily to contact the landlord. After some back and forth, a few days ago, he finally called and the landlord agreed to the sale for 260 million yuan in full.

Emily said that if they negotiated further, the price could be even lower, but Chen Yiran couldn't wait any longer. He was already eager to start renovating the house.

The process is simple. For cash purchases, there's no need to wait for bank approval; a 3% earnest money deposit is paid into a third-party escrow company's account. Next comes a home inspection. California law requires sellers to provide a disclosure report detailing any issues with the property. If the inspection is satisfactory, a final contract is signed, the escrow company handles the title transfer and title insurance, and then the transfer of ownership is complete.

Emily's voice was as lovely as ever when she called: "Buying with cash is quick. I've found you a custodian company in Palo Alto that specializes in high-value properties."

Finding a home inspection was easy; I Googled it and found the highest-rated one, scheduling it for the next morning. A bespectacled white man, armed with a flashlight and an infrared camera, spent three hours thoroughly inspecting the property. The report: structurally sound, very, very new.

After that, the procedures were followed, and Emily came over with the final contract. Chen Yiran wrote the check.

Emily smiled and said, "Congratulations, Mr. Chen, this house will be yours in a week."

Chen Yiran couldn't help but blurt out, "Sigh, now that the rental has been converted to residential, I won't be able to contact you if anything comes up."

"It's alright, I believe in Mr. Chen's abilities, and he will buy an even bigger house in the future."

Chen Yiran didn't respond. The house in this location was actually quite good, and there were quite a few billionaires around. Chen Yiran really liked the location and already had an idea for renovations.

After the house was settled, Chen Zelin also received good news.

"Mr. Chen, I've found a very suitable candidate. David Kao, Bachelor of Electrical Engineering from National Taiwan University, PhD from Berkeley, has worked at Cadence for a full fifteen years. He was a key participant in the core development of the QPlace placement engine."

"The direction is perfect."

"Yes, that's the direction you should be working on. He left Cadence in 2007 to work on a new algorithm, but the company said no, the existing product line couldn't be touched. After leaving, he taught VLSI at Santa Clara University and occasionally did consulting."

"Is he willing to come see me?"

"I was willing to meet him. To be honest, when I told him, he just said bullshit. Doing EDA at eighteen sounds like a joke to anyone. But I showed him that TechCrunch article, and he said he was willing to meet him."

After thinking for a moment, Chen Yiran finally asked, "Are you also from Tainan?"

Chen Yiran's question about Tainan was not without reason; Huang Renxun and Su Zifeng are both from Tainan. There are also many Taiwanese in other sectors of the chip industry. In short, there are a great many Taiwanese, especially from Tainan, among the Chinese Americans in the chip industry.

Chen Yiran wasn't sure if this would affect the recruitment process; it wasn't a political issue. Just like when he was dealing with parallel imports, Chen Yiran pretended to be from Changle, Fujian. Overseas Chinese don't really care if you're from China or not; they only treat fellow countrymen as family.

Therefore, Taiwanese-owned shops overseas naturally prefer to hire Taiwanese people. People from Fujian, on the other hand, prefer to apply for jobs at shops owned by people from Fujian.

Chen Yiran wondered if there was a similar connection in this high-end industry.

"Mr. Chen, don't worry, the people I found don't care much about this. They've all been out here for a long time, and many of them consider themselves Americans. Of course, it's easier to communicate with fellow Chinese. Although I didn't deliberately look for them, 70-80% of people in this industry are Chinese, regardless of their specific hometown, so they're all considered family."

"Okay. Let's meet at the same teahouse tomorrow afternoon."

The next day, Chen Yiran met Gao Dewei at the same Tianren Tea Shop.

Gao Dewei looks very young, and you can't tell that he is already over fifty. He wears thin-rimmed glasses, a plaid shirt and khaki pants, a very standard and stereotypical Silicon Valley engineer's outfit.

Chen Yiran had just taken Pidan to the park and came over without letting him go home. Pidan wagged his tail and sniffed Gao Dewei's pants, making the latter laugh happily.

After ordering drinks, the two began discussing business. Gao Dewei said directly, "Mr. Chen, I won't beat around the bush. If you want to work in EDA, I'm a professional, but you have to show me that it has potential. After all, even for a genius, entering this industry is a bit too far-fetched. I find it hard to imagine what new achievements you can make amidst the dominance of Synopsys, Cadence, and Mentor."

But if you can really convince me, I can say that I was one of the key contributors to Cadence's QPlace, and that Synopsys' IC Compiler is our biggest competitor, which I know very well. In this industry, there are only a handful of people worldwide who can build the core architecture of a placement engine, and I'm one of them. My understanding of EDA is certainly sufficient.

"Professor Gao, I don't know much about simulated annealing, and I've never written a force-directed algorithm before, but my approach is different."

"What's the entry point?"

"What is the essence of placement? It's about placing millions of components on a chip, satisfying connectivity and area constraints, and minimizing line length. This is an optimization problem. All traditional methods do the same thing: approximate the optimal solution using heuristic algorithms."

Gao Dewei nodded.

"However, the upper limit of heuristic algorithms is locked by human design. What if we take a different approach? Reinforcement learning, letting AI explore, try and fail, and learn in placement tasks. Don't give it preset rules, just give it signals, such as short lines, less congestion, and good timing. Let it learn the optimal strategy on its own."

"This is an interesting idea. But do you know how large the state space of a placement is? Millions of elements, each with its own position coordinates and connectivity. This state space is larger than the game of Go. Can current deep learning achieve that level of complexity?"

"Layered approach: first, perform global layout, then mesh the chip at a coarse-grained level. Then, fine-tune local details within this coarse-grained framework, reducing complexity. Furthermore, I've made breakthroughs within the deep learning framework: dirt label propagation, incremental analysis, and deep learning-driven layout optimization. These are all possible."

Gao Dewei picked up his water glass, took a sip, and asked, "What exactly do you need me to do?"

"You're essentially my technical partner, helping me build the placement team and develop an AI-driven placement engine from scratch. I need a model architecture for chip design problems, how to encode chip placement problems into a form that deep learning can handle. I also need sufficient domain knowledge, physical constraints, design rules, and timing constraints. As for funding, I have 1500 million in the company account, enough to last a long time."

"I want to make one thing clear: you're not just looking to make a quick buck and run, are you?"

Chen Yiran glanced around and said in a low voice, "You should know that my first project easily sold for 200 million. If I wanted to make quick money, there are plenty of projects better than this. My goals are much grander; EDA is just the first step. AI-driven end-to-end chip design, and the next step is chip architecture designed for AI."

"Okay, you've convinced me. But when we're hiring later, I want to join them; I don't want to work with idiots."

"Okay, I don't like working with idiots either. Let's leave it at that for now. I'll talk to the lawyer about the contract details, and we can sign the contract when we have the next interview."

"Um, there's a problem. Don't we have an office yet?"

"Oh, I forgot."

Gao Dewei said somewhat helplessly, "Can I take back my decision? I suddenly feel like our company is so unreliable. What do you mean, 'You forgot about the office'?"

"Well, I used to work alone, so working from home was fine. I actually think working from home is a major trend, maybe our company should just do it from home."

"Are you really not trying to save on office rent?"

"That's one of the considerations, and it's a bit difficult for me to go to work on time. Oh well, let's just rent one. I can work from home, and you guys can just clock in."

Gao Dewei was silent for a few seconds, then resignedly said, "Fine, you're the boss, you're the boss. You can't treat me unfairly, considering my annual salary and shares."

"How about an annual salary of 30 plus 5% equity?"

To be honest, the annual salary wasn't much, but Gao Dewei had never really started a business in his life, so 5% equity was enough to tempt him and make him leave his stable teaching job.

I'll give it a try anyway. If it doesn't work out, I can always go back to teaching.

Gao Dewei nodded, agreeing to the figure. He patted Pidan's head and said goodbye to Chen Yiran.

Chen Yiran stood at the door, watching the gray Prius drive away, and felt that stereotypes really do have some truth to them.

Having made so much money, do you still want to drive a Prius?

Chen Yiran got into his Cayenne and started to feel a little disgusted with it. Now he felt that the car didn't quite match his status.

It's unacceptable to leave no room for bullets.

Chen Yiran drove home and began designing renovation plans for the house. She also started thinking about a new car.

Tap the screen to use advanced tools Tip: You can use left and right keyboard keys to browse between chapters.

You'll Also Like