Tech Giants: Starting from Bankrupt Companies

261: Chapter 261 A Shocking Change in the World: The New Wave Surpasses the Old.Among all the international giants, SMIC’s attitude was the most unique.

Founder Zhang Rujing personally led the team. The 62-year-old industry titan wore a simple Zhongshan suit, followed by eight technical executives.

”Director Li, I’ll be direct,” Zhang Rujing said, not mincing words. ”SMIC’s 14nm FinFET yield is stuck at 65% and won’t go up, and the cost won’t come down. We hope to obtain the license for Xinghuo’s FD-SOI technology. The price... is negotiable.”

SMIC is the largest chip foundry in mainland China, but it has always lagged behind TSMC by 2-3 generations in advanced process nodes. The 14nm FinFET has been in development for four years, with an investment of over 10 billion rmb, yet the yield has never been able to break through, and the cost is 1.5 times that of TSMC.

Li Wanqing pondered, ”Elder Zhang, you are a senior in the industry, and we respect you. However, technology licensing involves national security reviews, and FD-SOI is our core asset...”

”We only want the process package; it doesn’t involve core IP or design tools,” Zhang Rujing said. ”Moreover, we can exchange it for our 28nm mature production lines—SMIC has 8 28nm lines with a total monthly capacity of 500,000 wafers. These lines can be entirely dedicated to manufacturing low-end chips for Xinghuo, allowing you to free up your high-end capacity.”

This was a heavy bargaining chip.

What Xinghuo currently lacks most is 28nm capacity—the online system sold out 400,000 units in 60 seconds, and the queued demand exceeds 2 million units. If they could use SMIC’s 500,000 units/month capacity, it would be equivalent to more than doubling their supply.

But technology licensing...

”I need to consult Director Su,” Li Wanqing said.

”Please do,” Zhang Rujing said calmly. ”We will wait.”

At 6:30 PM, Li Wanqing returned to headquarters with the final data.

In the conference center, Su Chen, Gu Xinghe, Lin Wan, and the core team of Xinghuo Chip were all present.

”Summary of today’s contracts and online orders.” Li Wanqing pulled up the data dashboard:

28nm chip:

On-site contracts: 3.2 million units

Online flash sales: 400,000 units

Total: 3.6 million units

Unit price: $2,800 ($3,360 for automotive grade)

Total amount: $10.08 billion

16nm chip:

On-site contracts: 1.05 million units

Unit price: $4,700 (discounted to $4,500 for large volume)

Total amount: $4.935 billion

14nm chip:

On-site contracts: 520,000 units

Unit price: $5,200

Total amount: $2.704 billion

NRE (Non-Recurring Engineering) fees:

Custom chip design fees: $32 million

Total:

Total wafers: 5.17 million units

Total amount: $17.739 billion

Equivalent to rmb: approximately 124.2 billion rmb (at a 1:7 exchange rate)

The conference room was quiet for a few seconds.

Gu Xinghe spoke first: ”One day... $17.7 billion? This is comparable to selling arms.”

”And this is only the first year,” Lin Wan added. ”Many are three-year framework agreements; the subsequent amounts will be even larger.”

Su Chen looked at the data: ”The domestic market digested 90% of the 28nm chips... This means that the low-to-mid-end chips for China’s electronics industry will basically be localized from now on.”

”Yes,” Li Wanqing nodded. ”More importantly, mobile phone manufacturers like Huawei, Xiaomi, and OV will fully switch to our 14nm/16nm for their mid-to-high-end models next year. Qualcomm and MediaTek’s market share in mainland China... might plummet from the current 70% to below 30%.”

”What about the international reaction?” Su Chen asked.

7:00 PM, TSMC Headquarters.

An emergency board meeting was underway.

The big screen showed a data comparison chart of Xinghuo Chip, especially the ”16nm vs 10nm” actual test chart.

Chairman Morris Chang, 87 years old, was semi-retired, but he attended the meeting in person today, his face as grim as iron.

”We underestimated their FD-SOI,” said Wei Zhejia, the Executive Vice President of Technology, his voice sounding dry. ”At the same 16nm, the power consumption is 35% lower, and the cost is 28% lower. If the data is true, our advantage in the mobile sector... is gone.”

CFO Ho Li-mei added: ”The more troublesome part is the price. They sell 28nm for $2,800, while our cost is $2,600. They can still have a profit at a low price, but we have shareholders who want profits and R&D that requires investment...”

”Are they dumping?” a board member asked.

”No,” Morris Chang said slowly, his voice old but clear. ”Their cost might really be only $2,200. The FD-SOI process steps are 20% fewer than FinFET, and the lithography layers are 15% fewer. This is a victory of the technical route, not dumping.”

He paused: ”Adjust the roadmap. Accelerate 3nm R&D, but more importantly... start FD-SOI technology reserves. We need to do FD-SOI too.”

”But... we have no accumulation in FD-SOI,” Wei Zhejia hesitated.

”Then buy it, or poach people,” Morris Chang said decisively. ”There must be people in Xinghuo’s FD-SOI team who were poached from GlobalFoundries and STMicroelectronics. Go approach them; poach them with double the salary.”

”What about the US side...”

”I will call the White House myself,” Morris Chang said. ”We cannot let them occupy the global market with low-priced chips. In addition, file an anti-dumping investigation application with the WTO—regardless of whether it works, delay them first.”

The meeting lasted until 8:00 PM.

Resolutions:

Accelerate 3nm R&D, target mass production in 2021

Start FD-SOI process R&D, target mass production in 2020

File anti-dumping investigation with the WTO

Poach Xinghuo’s FD-SOI team with double salary

But everyone knew these measures would be too slow to take effect.

Xinghuo’s technological advantage was at least two years ahead of TSMC.

And two years in the chip industry... is enough to change dynasties.

7:30 PM, Seoul, South Korea, Samsung Electronics City.

On the Samsung Electronics stock ticker, the numbers were sliding downwards.

Opening: 58,000 won per share

Closing: 51,000 won per share

Single-day decline: 12.1%

Market value evaporated: approximately 38 trillion won (about $35 billion)

Chairman Lee Jae-yong stood in front of the office window, looking at the reporters gathered downstairs, his face gloomy.

”Analyst reports say Xinghuo Chip will take away 30% of our Exynos processor customers,” President Kim Ki-nam said with his head lowered. ”And if Huawei and Xiaomi switch entirely to Xinghuo Chip, our mobile phone business in the Chinese market might... lose half its share.”

Samsung’s mobile phone market share in China has fallen from its peak of 20% to less than 3% now. If the chips are throttled again, they might really withdraw from this largest market in the world.

”What about our 7nm?” Lee Jae-yong asked.

”The yield is still hovering at 35%, and the cost is three times that of the 7nm claimed by Xinghuo,” the Chief Technology Officer said with a bitter voice. ”And... Xinghuo says they will mass-produce 7nm in the first quarter of next year, while our 7nm EUV (Extreme Ultraviolet lithography) version won’t be ready for mass production until 2020.”

Silence.

A long silence.

”Contact Qualcomm in the US,” Lee Jae-yong finally said. ”Propose joint development of the next-generation process to share R&D costs. If Samsung and Qualcomm join forces, perhaps we can still put up a fight.”

”And... what about Xinghuo?”

”Start the ’Reverse Engineering Project’,” Lee Jae-yong’s eyes were cold. ”Buy their chips, dismantle them, analyze them. At all costs, get the process details of the FD-SOI.”

”This violates business ethics and is also illegal...”

”Then do it secretly,” Lee Jae-yong interrupted. ”Set up an R&D center in China, hire Xinghuo’s former employees with high salaries. Use every means to get the technology.”