Recently, Intel CEO Chen Liwu gave an in-depth interview on a podcast to share his reform ideas, industrial layout and industry judgments since he took charge of Intel for 14 months.
Talking about why he chose to take on another heavy responsibility after becoming famous, Chen Liwu said there were several reasons. First, Intel is an iconic company that is extremely important to the industrial ecology and the United States; second, he wanted to decide to do another big thing after Cadence (EDA electronic design automation software company).
Chen Liwu said that a lot has happened in 14 months.
The first change is culture, clarifying the accountability mechanism and reducing layers of bureaucracy; the second is listening to customers and staying humble.
Third, all engineering teams report directly to it.
He revealed that Intel is formulating a clear roadmap and vision for the next five to ten years.
He pointed out that Intel used to be an old-school company that relied on spreadsheets, but now it is transforming it into an AI-enabled enterprise – not just in the design process, but in the entire organization to fully embrace AI and reduce reliance on spreadsheets.
When talking about the company’s major business sectors, Chen Liwu used the metaphor of “crawling – walking – running”: The company is still crawling in the past few months, but people have begun to see the potential.
In terms of products, Intel still has market share in PC clients, but it must significantly improve performance.
He himself is quietly building CPU architecture, GPU architecture and software architecture teams to prepare for leapfrog leadership.
In terms of foundry business, he admitted that there is still a big gap with TSMC, and now he is focusing on laying a good foundation.
But by 2030 to 2032, people will start to see the true potential of Intel.
Chen Liwu also said that his intuition as a VC told him to look for opportunities with 10 times returns.
At Cadence, he brought investors a 76-fold return; Intel is larger and more difficult to copy, but his goal is to achieve a 10-fold return within 5 to 10 years.
Chen Liwu enjoys a high personal reputation in the semiconductor industry, while Intel fell into a trough a year ago.
Talking about why he took over this hot potato, Chen Liwu admitted that most people around him advised him to retire, but two core reasons prompted him to join: “First, this is an iconic company, which is extremely important to the entire semiconductor ecosystem and the United States; second, after Cadence, I decided to do another big thing.” He bluntly said that his core original intention of taking over the job was to “save Intel.”
Early on taking office, Chen Liwu encountered a sudden crisis. U.S. President Trump asked Chen Liwu to resign on the grounds of conflict of interest.
When Chen Liwu recalled this unexpected incident, he said: “I convinced myself at the time: I didn’t need this job. I did this purely to save Intel.” After two interviews, he told President Trump about his experience of being born in Malaysia, growing up in Singapore, graduating from MIT, and living in the United States for a long time, and finally got the opportunity to continue to perform his duties.
When talking about internal changes and long-term planning since taking office, Chen Liwu proposed a three-step development framework of “crawl-walk-run” to comprehensively transform Intel’s original system and culture.
Chen Liwu said that the first step is to change the culture, clarify the accountability mechanism, and speed up decision-making. “I am used to the pace of a startup company, where everything moves at the speed of light, but Intel has layers and layers of bureaucratic meeting systems. This is what I must change.”
He also mentioned a number of key measures: requiring all engineering teams to report directly to him and personally mastering front-line technical issues; insisting on completing all talent recruitment in person without resorting to headhunters; promoting enterprises to fully embrace AI and eliminating the old management model that relied on spreadsheets in the past; and adjusting the age structure of the team to simultaneously absorb senior technical experts and young AI talents.
When Chen Liwu took over Intel, the outside world had different views on Intel’s foundry business. Some believed that it needed to exit, thinking that it was too expensive and unworkable.
But Chen Liwu finally judged that OEM is extremely important to the United States and the entire industry.
“We have all experienced supply chain challenges. Any major semiconductor company must seriously think about supply chain issues. It must have a robust and resilient supply chain and cannot completely rely on one or two geographically concentrated suppliers. More and more people will realize that manufacturing in the United States is crucial.” Chen Liwu said.
He revealed that Intel’s most advanced process, such as 18A, is 1.4 nanometer level, and 1 nanometer and 0.7 nanometer are already being planned.
The process nodes are getting smaller and smaller, the line width is thinner than a hair, and the complexity is extremely high. If any step is wrong, all previous efforts may be wasted.
Because of this, the requirements for manufacturing accuracy are getting higher and higher, which will increasingly become a bottleneck.
Chen Liwu also admitted that in terms of foundry, there is still a big gap between Intel and TSMC. They must remain humble and focus on laying a good foundation – IP, yield, defect density, cycle time – to make foundry more efficient and reliable.
“This is a trust business. Customers must trust you before handing over wafers to you. These things will take longer, but I think that by 2030 to 2032, people will begin to see the real potential of Intel.” Chen Liwu said that he respects TSMC very much and is a good partner with TSMC, and the industry needs more production capacity to serve customers.
Chen Liwu revealed that the Terafab project promoted by Intel and Elon Musk stemmed from the mutual judgment of both parties that the development of semiconductor infrastructure lags behind the growth of AI demand in terms of production capacity, production efficiency and power consumption efficiency.
Under the framework of this cooperation, Musk decided to build his own wafer factory, and Intel will provide technical and process support to assist him in accelerating production.
Chen Liwu said that he holds weekly meetings with Musk’s team and the cooperation is progressing smoothly.
He also mentioned that Musk has unconventional ideas at the operational level, such as discussing whether to allow smoking in certain areas of the clean room. “I may not go that far, but it may be possible in certain areas. The key is to keep an open mind.”
As the chip manufacturing process advances, more and more people are discussing the physical limits of chip shrinkage, and the line width is too narrow to continue shrinking.
In this regard, Chen Liwu said that this road will become more and more expensive and difficult, and it requires industry chain partners to work together to promote the improvement of yield and performance.
He also specifically mentioned the help of advanced packaging and materials in cracking the physical limits of chips.
He said that TSMC has CoWoS and Intel has a next-generation solution called EMIB, which must ensure that it can achieve the yield required by customers in the mass production stage.
Chen Liwu said that when traditional miniaturization began to encounter bottlenecks, he began to return to the material level to find breakthroughs – gallium nitride, silicon carbide, and indium phosphide. He has investments in these three directions.
In terms of packaging materials, he began to pay attention to glass – glass is a good heat dissipation and insulating material, and he invested in a company called 3DGS.
He also revealed that Intel has about 1,000 patents on modules, and how to integrate substrates and modules is an important issue.
Recently, Intel also announced advanced packaging manufacturing cooperation projects in India and New Mexico, USA.
In addition, he is also looking at synthetic diamonds, another excellent insulation material, and has invested in a diamond wafer company.
Chen Liwu said that the spirit of engineers is like this – you constantly encounter bottlenecks, and then find ways to overcome it or bypass it.
Chen Liwu said that the PC client is Intel’s basic disk, but Intel is extending to the edge, to physical AI and intelligent AI.
“In the past, you only provided servers and PCs to humans, but now there is a whole new dimension – millions of agents need access to computing power and software stacks. I think Intel has opportunities in both the directions of agent AI and physical AI, and this game is not over yet,” said Chen Liwu.
As a successful long-term investor, Chen Liwu admitted that venture capital and entrepreneurship is in his blood and he enjoys it very much. He has invested in 159 company IPOs and 126 M&A exits, including more than 200 semiconductor investments, 38% of which were in the United States.
He believes that in addition to agent AI, physical AI is the next major frontier, and we must carefully look at the full stack. For this reason, he is still deeply involved in many investments related to cutting-edge models. “I am very optimistic about open source cutting-edge technology for physical AI. This is a gold mine.”
When talking about Intel’s long-term goals for the future, Chen Liwu said that his instinct for venture capital is to look for opportunities with a 10-fold return.
At Cadence, he went from acting CEO to retirement, and the stock price rose from $2.40 to a return of about 76 times for shareholders; when the executive chairman expired, it was about 85 times.
Intel is larger and more difficult to copy, but his goal is 10 times – to achieve a 10 times return within 5 to 10 years. As a VC at heart, this is his goal.
