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

TSMC, as a key partner of Nvidia, actually paid attention to Chen Yiran quite early. Although Chen Yiran was turned away when he inquired about PDK certification, that's standard procedure at TSMC.

Several Taiwanese OEM companies have always adhered to a client-centric corporate culture. Naturally, when they themselves are the clients, they tend to become somewhat arrogant.

Moreover, TSMC is the largest customer of the three major EDA companies and has enjoyed preferential treatment from them for many years. Therefore, TSMC has no interest in other smaller companies at all. The so-called 9 to 15 month scheduling is actually more like discouraging them from entering the market.

It means, take a good look at yourself in the mirror before you start working with us.

Later, Nvidia and Chen Yiran began to collaborate, and TSMC was aware of the emergence of DeepFlow, but they still didn't think there was much of a problem.

Their original plan was to see what results Chen Yiran had produced at this meeting, and if they were suitable, they would consider agreeing to cooperate.

However, Chen Yiran's various technical demonstrations on stage shocked Rick Cassidy, the vice president of TSMC, who was in the audience.

Cassidy is a pure manager with a law degree. But that doesn't mean he doesn't understand technology. He has worked incredibly hard to get to his current position, and ten years of honing his skills have given him a very keen technical intuition.

He sensed that Chen Yiran's technology was revolutionary. Although there was no systematic calculation or evaluation, he just felt that Chen Yiran's technology was not just the kind of technology that could challenge EDA, as everyone said.

He felt this was a revolutionary technology, one that could even lay the foundation for the next decade.

However, he wasn't panicked yet. He didn't think Chen Yiran would refuse to cooperate because of a minor unpleasantness. He just needed to be more polite this time.

But after he found Chen Yi and said three sentences, the other party left.

Cassidy wanted to catch up and try again, but the other party was followed by more than a dozen employees. Perhaps because the company had achieved success, these people were all extremely excited, shaking their heads and looking like small-time thugs from the Chicago mob.

Cassidy felt that if he chased after them, he might even get attacked, so he decided to ignore it and report back to Taiwan first.

He needs a higher level of authorization and better terms.

However, the feedback from Taiwan was not very strong. The technical team was taking a very long time to assemble, and they did not provide a specific technical report even after two days.

Chiang Shang-yi, the co-COO in charge of R&D, was also slow to react. He didn't even look at the US report in detail and just told Cassidy to give more money and have a proper talk.

If that doesn't work, TSMC can offer a 20% discount during PDK certification.

Chiang Shang-yi's confident response led Cassidy to misjudge the situation. He suddenly felt confident again. After all, TSMC had many customers, and Cassidy didn't believe anyone would actually give up. He thought Chen Yi-ran's reaction was just a negotiation tactic.

However, on the morning of the last day of the conference, Nvidia staff distributed flyers announcing that due to Chen Yiran's recent presentation, especially regarding manufacturing EDA, and the excessively exaggerated data from the Tiya (the little white rabbit) AI, a separate manufacturing mini-summit was being organized after the closing ceremony. Those interested could stay and listen.

Samsung, GlobalFoundries, SMIC – this lineup sent chills down Cassidy's spine.

They are indeed the industry leaders, but they are definitely not far behind Samsung, plus GlobalFoundries, which is in third place.

Anyone with a discerning eye can see that this is the beginnings of a new alliance, an alliance that uses new technologies.

Cassidy immediately decided to go to hell with the authorization from headquarters. Those people in Taiwan were too slow to react. In Chinese terms, "a general in the field may disobey the emperor's orders."

He's planning to do something big, starting with a high price: five million, or even ten million a year, or even twenty million.

Let EDA resolve its own love-hate relationships; TSMC is different from other foundries.

They absolutely cannot get involved in the war between EDA companies, as this is completely incompatible with TSMC's long-standing corporate culture. If they get involved in the war between their customers, they are likely to lose TSMC's most important advantage: its pure foundry nature.

At that point, they'll be no different from Samsung, and people will choose Samsung over them because of competition on other issues. Of course, they might also choose TSMC because of its deep involvement in the EDA wars.

Cassidy approached Chen Yiran again, but this time the result was even worse.

He had no choice but to contact Taiwan again as soon as possible, and this time his attitude was very firm, saying directly that this was a matter of life and death.

He spoke privately with Chiang Shang-yi for half an hour, explaining all the technical aspects of the situation to him.

Jiang Shangyi holds a Master's degree in Electrical Engineering from Princeton University and a Ph.D. in Electrical Engineering from Stanford University.

He had retired in 2006, but was brought back by Morris Chang in 2009. As co-chief operating officer, he was in charge of research and development.

As an absolute technical genius, Jiang Shangyi felt a little flustered after truly understanding the skills Chen Yiran demonstrated.

He is currently working on advanced packaging, a field that no one has ever ventured into before. It is unknown how much impact Tiya will have on the manufacturing end, but it seems that he can perform finite element thermal simulation.

What if he spends a lot of time and effort researching but still can't figure it out, and then a year later, the other side easily gets it working?

Chiang Shang-yi immediately took the lead in setting up a technical evaluation team. This time, there was no perfunctory work; the heads of each department and key technical personnel were all brought in.

The intelligence work between wafer foundries is actually very meticulous. After Samsung received Apple's order, TSMC had a team specifically monitoring Samsung for many years.

What equipment did Samsung buy, how many units did they buy, and which design company recently had dinner with Samsung representatives?

They would also buy Samsung chips from the market, disassemble them, scan and photograph them with an electron microscope, and analyze them layer by layer.

They even hire outside help; there are several chip teardown and analysis companies that specialize in providing this service.

Based on various published information from the United States and the data at hand, Chiang Shang-yi convened a meeting at 11 p.m.

Morris Chang also attended in person, and various unfavorable data were displayed on the large screen in the conference room.

Morris Chang covered his face and said, "So, you mean that after we finally improved the mass production yield of 40nm from 68% to 80%, we'll immediately face Samsung, who will improve from 85% to 90%?"

A department head stood up, found a diagram on the table, and handed it to Morris Chang, saying, "It's like this, the Tiya's control over critical dimensions is too precise. With fewer hot spots, the lithography window is wider, naturally resulting in a higher yield."

We've done a rough assessment, and conservatively speaking, Tiya can reliably improve the yield rate by about 5%.

This isn't actually the most frightening aspect; what's truly alarming is its iteration speed. This could result in us starting ten percentage points behind in the future 28nm race.

Morris Chang glanced at Chiang Shang-yi, his meaning clear: he hoped this leading figure in TSMC technology could provide some good news.

Jiang Shangyi thought for a moment and said, "I have good news. Chen Yiran's EDA should be able to develop some auxiliary functions for packaging. If we can talk to him, our advanced packaging may be able to be realized ahead of schedule."

Morris Chang glared at Chiang Shang-yi, then looked around.

Some people in the conference room looked down at their shoes, some looked up at the ceiling, and a few who had been working all night simply pretended to be unable to stay awake and closed their eyes to sleep.

Morris Chang sighed and said, "It's outrageous. I'm old and I still have to come out and struggle, patching up your problems. If I had known, I wouldn't have come back. It would have been so much better to retire at home. Sigh, so what happened before? How did I talk to them? How could I have made them so angry?"

Everyone remained silent. Zhang Zhongmou ignored the others and stared intently at Jiang Shangyi.

Chiang Shang-yi immediately panicked, stemming from his fear of his seniors. When he first started working, as an intern, Morris Chang was already his superior.

Jiang Shangyi coughed lightly and replied, "Ahem, well, actually, nothing much was said. Everything was done according to normal procedures, but who knew that young man would be so aggressive. What do we do now?"

Morris Chang stood up abruptly: "What can we do? What else can we do but swallow our pride and apologize? You'll come with me. I'll be 80 next year, so someone has to share the shame with me."

"No, I still have a lot of work to do."

"Stop talking nonsense and let's go, hurry up. The tech team continues testing. Find someone to evaluate the potential of DeepFlow. We're not experts in this, so we need to find someone to help. I need your report when I get to the US."

After everyone had left, Jiang Shangyi asked with some worry, "I have a bad feeling about this. I didn't mention it in the previous meeting. If we go over there now and apologize, wouldn't that mean we're on his side? Will the three major EDA firms have any objections? After all, they've always given us priority, and Samsung always gets updates later than us."

Morris Chang was silent for a moment, then sighed: "I've thought about it carefully, and this matter probably can't be resolved peacefully. My old face doesn't matter, it's just an apology. But if he pressures our team, then it will indeed be a bit troublesome."

Samsung is a follower, so of course they can take big gambles. Should we gamble too, betting that little kid can really succeed? Besides, Samsung can afford to gamble, but we can't.

I'd still prefer to see if he can just apologize and let it go, without making a big fuss. If he's really stubborn and uncooperative, then we can go find Apple on the way.

Chiang Shang-yi, an honest man who only sees technology, was a bit bewildered: "Find Apple? Go and snatch orders now?"

Morris Chang said decisively, "Yes, we must secure orders now! If we can't reach an agreement with Chen Yiran, we need to be prepared if that EDA business really takes off and ultimately makes us less competitive. In that case, we absolutely must get Apple's orders."

------

TSMC will be keeping an eye on Samsung, and Samsung will certainly be keeping an eye on TSMC.

Since they were the ones initiating the attack, their objective was clear: they specifically sent several employees on a company-funded trip to keep an eye on the airport.

After discovering that TSMC's private jet was making moves and flying to the United States, they immediately took action and began contacting the media.

Samsung had been preparing for this for two weeks and was determined to make it a huge success.