Video Title: Re-engineering the Semiconductor Supply Chain with Intel CEO Lip Bu Tan
Video Author: No Priors
Compiled by: Peggy
Editor's Note: Against the backdrop of sustained heating investment in AI infrastructure, the semiconductor industry discussion is shifting from "Is GPU supply sufficient?" to "Can the entire computing and manufacturing system support the next phase of AI expansion?" Over the past two years, the market has focused more on models, compute clusters, and the NVIDIA ecosystem; but as long-term AI demand growth gradually becomes consensus, a more critical question emerges: If chips, packaging, power, materials, memory, and manufacturing capacity all become bottlenecks simultaneously, what kind of new semiconductor supply chain does the AI industry truly need?
This episode of "No Priors" invites Intel CEO Lip Bu Tan to discuss Intel's transformation, U.S. domestic manufacturing, foundry business, AI's renewed pull on CPU demand, and new manufacturing collaborations like TerraFab. Lip Bu Tan is both a long-term semiconductor investor and an industry operator from Cadence to Intel, so the value of this conversation lies not in presenting a single company narrative, but in showing how an industry expert re-understands the semiconductor structure in the AI era.
In this conversation, Lip Bu Tan deconstructs "how to revive Intel" into a set of more fundamental structural questions: how to repair the balance sheet, how to refocus the product line, whether advanced manufacturing can return to the U.S., whether AI workloads will redefine CPU value, and how semiconductor investment should revolve around real bottlenecks.
First, Intel's problem is shifting from "product lag" to "organizational and capital structure reconstruction." In the past, external discussions about Intel often focused on process node stumbles, GPU absence, and insufficient foundry competitiveness. But what Lip Bu Tan emphasizes first is not a specific generation of products, but the balance sheet, organizational culture, and customer trust. The path he proposes is to "crawl," then "walk," and finally "run": first strengthen the financial foundation, simplify the product line, bring the engineering team closer to the CEO and customers, and then gradually rebuild the roadmap. This means Intel's revival cannot be completed by a single new product launch, but is a systemic repair of organizational speed, capital patience, and technology roadmap.
Second, AI's demand for computing structures is becoming more complex. Previously, the AI narrative was almost dominated by GPUs, with training clusters becoming the clearest consensus in capital markets. But Lip Bu Tan points out that with the development of Agentic AI, reinforcement learning, multi-agent orchestration, and edge computing, CPUs are becoming important again. The CPU-to-GPU ratio may shift from 1:8 in the training era to 1:4, or even close to 1:1 in some scenarios. This means AI infrastructure will not have just one chip winner; future competition will revolve more around system-level combinations for different workloads: CPUs, GPUs, NPUs, advanced packaging, software stacks, and foundry capabilities will all become part of the same computing network.
Third, semiconductor manufacturing is transforming from a commercial efficiency issue back into a national infrastructure issue. Over the past three decades, global chip manufacturing became highly specialized driven by efficiency, with advanced manufacturing capabilities concentrated in a few regions and a few companies. But supply chain shocks, AI capacity demand, and geopolitical risks make "relying on players in just one or two geographic regions" increasingly unsustainable. Lip Bu Tan compares the U.S. government's stake in Intel to the early relationship between TSMC and the Taiwan regional government, pointing to a new industrial policy consensus: for capital-intensive, long-cycle, strategically critical manufacturing systems, governments, sovereign funds, and long-term capital will re-emerge as key participants.
Fourth, the logic of semiconductor investment is shifting from "betting on hot sectors" to "finding real bottlenecks." The keyword Lip Bu Tan repeatedly mentions is not valuation, but bottleneck: interconnect, photonics, EDA, advanced packaging, power conversion, heat dissipation, new materials, memory, helium, electricity—all could become constraints in the AI expansion process. In the past, semiconductor investment was avoided by VCs due to high capital expenditure, long tape-out cycles, and high customer switching costs; now, as AI demand pushes these bottlenecks to the forefront, semiconductors are once again becoming an area of common concern for venture capital, strategic capital, and industrial capital. This means truly valuable investment is not simply chasing "AI concepts," but judging which link is becoming the constraint for the next round of expansion.
Fifth, future computing will not only exist in hyperscale data centers. The past SaaS and cloud computing era formed a highly centralized computing paradigm, but robotics, defense, home devices, Physical AI, and Agentic AI are making on-device and edge computing important again. Lip Bu Tan does not deny the continued expansion of large AI data centers, but he is more concerned about what applications these infrastructures ultimately serve. In other words, computing power construction can only generate long-term value when combined with sustainable, large-scale applications. This also means the next stage of AI competition is not just about "who builds more data centers," but "who can connect computing power, chips, and application scenarios into a scalable system."
If this conversation is condensed into a single judgment, it is: AI is pushing semiconductors from single-chip competition towards a comprehensive restructuring of the supply chain, capital structure, manufacturing capabilities, and system architecture. In this sense, the subject of this article is no longer just whether Intel can revive, but whether the computing infrastructure of the AI era needs to be redesigned from the ground up.
The following is the original content (slightly reorganized for readability):
TL;DR
· AI's bottleneck is no longer just GPUs, but an industrial system constraint composed of power, memory, packaging, materials, and manufacturing capacity.
· The key to Intel's revival is not a single-point product counterattack, but a systemic repair of the balance sheet, engineering culture, customer trust, and product roadmap.
· CPUs are becoming important again, not because the GPU narrative is cooling, but because Agentic AI, reinforcement learning, and multi-agent orchestration are creating new computing load structures.
· Semiconductor foundry is not just a manufacturing business, but a trust business; before customers deliver wafers, they must first trust that yield, cycle time, and reliability will not destroy their revenue.
· The signal of TerraFab is that AI demand growth is so fast that leading customers are starting to intervene upstream in manufacturing infrastructure, rather than passively waiting for chip supply.
· Rebuilding advanced chip manufacturing in the U.S. relies not just on factory subsidies, but on a recombination of government capital, long-term funds, industrial customers, and manufacturing capabilities.
· The core of semiconductor investment is not chasing hot concepts, but identifying the bottlenecks that truly limit industry expansion, such as interconnect, power consumption, heat dissipation, packaging, and new materials.
· Future AI competition will not only occur in hyperscale data centers; edge devices, robotics, defense, and Physical AI will push computing power back to the application site.
Original Compilation
Host:
Hello everyone, welcome back to No Priors. Today, Elad and I are joined by Lip Bu Tan. He previously worked at Walden, later served as CEO of Cadence, and is now the CEO of Intel. We talked about his plan to transform Intel, the U.S. government becoming a significant Intel shareholder, how to be a great semiconductor investor, and whether we can actually manufacture chips in America. Welcome, Lip Bu Tan.
Why did Lip Bu Tan take over Intel?
Host:
Lip Bu Tan, great to see you. Let's start with the most direct question: Intel is an extremely important American semiconductor company, but the CEO role is incredibly difficult. Why did you still take this job?
Lip Bu Tan:
That's a good question. I am 66 years old this year. Many people would say, you should retire, not take on the hardest job in this industry. I did this for several reasons. First, Intel is an iconic company. It is very important to the semiconductor ecosystem and very important to the United States. So I decided, after Cadence, to do one last thing.
Host:
A lot has happened in the past year. What surprised you the most?
Lip Bu Tan:
What surprised me the most was something my previous work experience and training never taught me: one day, early in the morning, President Trump asked me to resign, citing a conflict of interest with no exceptions.
So I first had to convince myself: first, I don't need this job. I took it purely to save Intel. Therefore, I put personal issues aside first, and then thought about what I could do to help Intel.
The good news is, I arranged a meeting on Thursday morning and met him on Monday. He was willing to listen to my explanation. I told him I was born in Malaysia, grew up in Singapore, later went to MIT, and then lived in the U.S. for a long time. I have never lived in a country other than the United States.
I told him all of this, he listened very carefully and gave me a chance. So I am very happy.
Host:
Now you have the opportunity to really get to work. You just said the goal of this job is to "save Intel." In your view, what does it specifically mean for Intel to win again and prosper again?
Lip Bu Tan:
I have been in office for 14 months now. A lot has happened in these 14 months.
First, is changing the culture. It was clear we needed a stronger sense of accountability. Second, decision-making must be faster. I am very accustomed to startup culture: moving at the speed of light, no bureaucracy, no layers upon layers of meetings.
So the changes I pushed include: strengthening accountability, listening to customers, and making customers happy. Some customers would say, Lip Bu Tan is humble, willing to listen, and willing to solve the problems they face, striving to satisfy customers.
Also, from day one, I decided to have all engineers report directly to me. I am an engineer by training; I want to know exactly where the problems are and what needs to be corrected. I want to listen to customers, satisfy them, and ensure we have the right products, simplify the product line, and establish a clear roadmap and vision for the next five to ten years.
Intel's Ten-Year Vision: First Save the Balance Sheet, Then Rebuild Products
Host:
What is your vision for Intel over the next decade?
Lip Bu Tan:
I think there are several things. First, whether at Cadence or at Intel, I have always believed: first learn to crawl, then stay humble, listen to customers; second step, start walking; finally, start running and sprinting. This is my culture: step by step.
For me, the first step is to strengthen the balance sheet. Intel's balance sheet was, to some extent, very poor. So I am very happy to see the U.S. government become a significant shareholder.
Just as I explained to President Trump, when TSMC started, it also had the Taiwan regional government as a shareholder. Look at Japan, Singapore—semiconductors are fundamentally infrastructure, and the U.S. government needs to provide support.
Second, I am also very happy that my old friend Jensen Huang invested 5 billion dollars to support me. I am glad I did at least some things right. That 5 billion he invested has now become 25 billion, or even more.
Also, there is Masayoshi Son from SoftBank. I used to serve on the SoftBank board, and he also reached out to help me. So we first strengthen the balance sheet, then focus on products. I significantly simplified the product line, listened to customers, and pushed for next-generation leadership products.
To some extent, we are also lucky. Now with the rise of Agentic AI, CPUs are in very high demand. In the past, in training scenarios, the CPU to GPU ratio might have been 1:8; now I see it potentially becoming 1:4, or even 1:1. I am very happy that CPUs are becoming important again.
I talked to some AI model developers. They said that for reinforcement learning, and for the speed of orchestrating a large number of agents, CPUs actually have an advantage. So in a way, I am very happy that the market demand for my CPUs is very high right now.
Overall, we need to push hard on the product side, especially in data center servers. Another part is our wafer foundry business.
Initially, this is a capital-intensive business, not easy. You need to have several conditions. You need to have the right IP to support customers. For example, if a customer is doing mobile-related products, you must have low-power IP. Without these capabilities, you cannot serve them.
Foundry is a service business, and also a trust business. If a customer places an order with you, gives you their wafers, but the yield is not good, their revenue will be damaged, or they might even miss opportunities.
So, for us, it is very important to focus on yield, defect density, cycle time, and ensure we can meet customer needs and serve customers with high quality and reliability. These are the things I am truly focused on.
Ultimately, you also have to move towards the full stack. Not just silicon, but also software. Some customers will directly ask me: can you give me a full rack system? This means you have to build systems. So we are advancing these things step by step, quietly, and recruiting the best talent possible.
By the way, I do all the recruiting personally, without using headhunting firms. So sometimes, having a strong network list and knowing who to call is very helpful.
Host:
You've been in this industry for a long time. You were previously the CEO of Cadence, for about 12 years, I remember?
Lip Bu Tan:
13 years.
Host:
13 years. And then two years as Executive Chairman, so 15 years in total.
Lip Bu Tan:
At the time, I initially only agreed to do it for three months.
Host:
Three months?
Lip Bu Tan:
Right. So I'm very careful now. Once you say "I'll only do it for three months," it could end up being 15 years.
What is TerraFab? Why does Musk want to build his own fab?
Host:
It looks like you still have a long road ahead at Intel. Another widely discussed big project is TerraFab and your collaboration with Elon·Musk. Can you tell us more about how this project came together? What is your involvement? How do you collaborate?
Lip-Bu Tan:
Of course. Elon·Musk, I think we all agree, is one of the best entrepreneurs of this century, perhaps even the best. We share a common judgment: semiconductor infrastructure has not actually kept pace with the growth of AI. You need capacity, you need productivity, and you need efficiency. These are the problems he and I both see: there is indeed a missing link here.
Second, I'm very happy to work with him. He is very unconventional. I call it "non-traditional." He questions every step: why do things the traditional way? In a way, it's very refreshing. I like that. I like working with people who have different perspectives, and then together we figure out the best path. Both sides learn a lot in the process. Obviously, he also has his own vision: his robots, his cars, all require massive amounts of silicon wafers.
Host:
Can you explain what TerraFab is? Many people might not be familiar with it.
Lip-Bu Tan:
TerraFab is his decision to build his own wafer fab. At the same time, we are happy to collaborate with him, ensuring we can work together to get him into production faster, achieve volume production sooner, and use some of our technologies and processes. This is something we are working on together. His team is excellent, and I communicate with them weekly. Working with them is very exciting.
Host:
He has also mentioned some ideas, like wanting to smoke in the cleanroom, things that are usually considered……
Lip-Bu Tan:
Right, right, and hamburgers. I don't think I'll go that far. Maybe some areas of the cleanroom could do it. But the key is to keep an open mind. We also listen and see what things are feasible.
How is AI reshaping the global semiconductor supply chain?
Host:
It's truly exciting to see you reshaping this company in the U.S.: on one hand, gradually building the foundry business, and on the other, collaborating on projects like TerraFab. Looking at the global AI and semiconductor supply chain, that is, if you observe how AI is reshaping the supply chain in a macro way, by country, you'll find different countries are affected differently.
For example, regarding claims that AI causes layoffs, I think most are currently exaggerated. Many layoffs are really just due to over-hiring during the 2020 pandemic. But what I see being cut first are often outsourcing companies, because businesses prefer to cut external manpower before internal employees. So they cut external customer service and external IT. This impacts some countries with large BPO industries more, like the Philippines, India, etc. They might be hit by AI in the short term.
If you ask further how companies in various countries can positively participate in the future of AI, it almost requires a country-by-country analysis. Places with cheap energy can host data centers; places with the capability to train models can train models, but perhaps only the U.S. and one or two other places have that capability.
How do you view the changes in the global semiconductor industry supply chain? Should certain countries invest more? For example, Israel has a presence with Mellanox, Nvidia, and Intel—should it do more in semiconductors? Should the Philippines return to a manufacturing base? How do you think about these issues from a global perspective?
Lip-Bu Tan:
That's a good question. Clearly, AI is changing the entire landscape. I think its impact will be bigger than the internet, and more profound. AI can initially help you do things more efficiently. Many agents can help you complete tasks that were originally tedious but necessary, and do them faster. So it can significantly improve efficiency. Even in semiconductor design, AI can enhance efficiency, for instance, in timing—how quickly a design can be completed; second is cost. So these will help companies improve efficiency.
There are also several bottlenecks in AI demand and growth. First, of course, is the power constraint everyone knows about. Some countries simply don't have enough power, so they will be affected. Second, many people don't realize that helium's impact on the semiconductor industry could also be very significant. Third, everyone knows there is a severe memory shortage right now; everyone is scrambling for memory. Even if you want to build a fab to increase capacity, it takes several years. The same goes for CPUs and GPUs—demand will be very high. Prices will also rise because we have to pass the costs on to customers. So all these will affect the growth of the entire industry.
Overall, I think the companies hit hardest are those that haven't embraced AI. Because AI can help businesses improve efficiency across various functions. We should embrace AI and find better ways to use it, whether for prediction, design, or various workloads. There is huge potential here.
Host:
Many people's simple objection to whether TerraFab or Intel's foundry business can be competitive actually centers on one question: some factors are inside the fab, like the IP you mentioned, the speed of business operations; there are also external factors. Elad also talked a lot about this earlier.
One of them is labor cost, and actual manufacturing capability. By investing in the foundry business, you clearly believe there is a possibility: we can manufacture locally. Elon believes this too. Can you talk about this issue? How real is this constraint?
Lip-Bu Tan:
You mean the labor constraint?
Host:
Yes.
Lip-Bu Tan:
When I was deciding whether to double down on the foundry business or exit it, there were many voices in the market. As you've seen, many people said it's too expensive, it won't succeed, it won't succeed. But I ultimately decided that this is very important for the United States and very important for the entire industry.
We have all experienced supply chain challenges. For any major semiconductor company, you must seriously consider the supply chain issue. You need a robust and resilient supply chain; you cannot rely on just one or two players located in different geographical regions.
So I think more and more people will realize that manufacturing in the U.S. is crucial. And the most advanced processes, like our 14A, which is about 1.4 nanometers, we have already started planning for 1 nanometer and 0.7 nanometers. The dimensions are getting smaller and smaller, even much finer than a hair. So the complexity is very high, and it's not easy to do. If any step goes wrong, all previous efforts are wasted. So manufacturing must be very precise.
From this perspective, this will increasingly become a bottleneck. We have great respect for TSMC; it is a great partner. More importantly, both sides need more capacity to serve customers. So we decided to grit our teeth and invest for the long term. I think in the long run, this is very critical, and it's where I can create more value for the industry.
Host:
People have long been discussing that eventually, one day we will hit a resolution limit and can't shrink any further. The line widths will become too small to continue advancing. When do you think we will actually hit this limit?
Lip-Bu Tan:
That's a good question. I think now we have 18A, next 14A will go into mass production, and I can still see the path to 10A and 7A. So I think this road can continue. But it will become increasingly expensive and difficult. That's why we need partners. We can't do it alone. We need to collaborate with material suppliers and equipment manufacturers to ensure we truly improve yield and performance.
Another part is also becoming a bottleneck, and that is advanced packaging. Everyone knows about TSMC's CoWoS. Now we also have a very good next-generation solution called EMIB. I must ensure it can achieve mass production with production yields that meet customer requirements.
Now Moore's Law is also starting to lose momentum, as you said. So I'm also researching some new materials, going back to the material level, back to the periodic table. I invested in three types of materials: gallium nitride, silicon carbide, and indium phosphide, and I'm observing how these new materials can drive the next step forward.
In packaging, I started investing in glass. Glass is a very good thermal insulator, so I invested in a startup called 3DGS. Later I realized that Intel has about a thousand patterns on a module, so how the substrate and module are combined is very important.
We just announced a large project with the Indian government for manufacturing in India and New Mexico in the U.S. So advanced packaging is very important. I also started looking into lab-grown diamonds. They are also very good insulating materials. So I also invested in Diamond Foundry. These are all directions worth watching for the next generation. That is to say, new materials, new substrate materials, and new design methodologies will all push the industry forward.
As an engineer, you always hit walls. But after hitting a wall, you either find a way to jump over it or go around it, ultimately achieving a better result. As someone who has invested in semiconductors for a long time and also participated in building the semiconductor industry, from EDA tools, to design, to manufacturing, having this experience is actually very helpful. Now I can use my own way to make a small contribution to the industry.
The Key to Semiconductor Investment
Host:
What you just said is very interesting: there are always things you can go around, but physical limits do exist. When you reach a scale like 7 angstroms, you hit constraints and must look for new materials or other detour paths.
The interesting question is, we've been discussing this topic for a long time. I remember 20 years ago, people were saying we would eventually reach a point where there's no space left on the chip. Will you encounter some kind of asymptote that levels the performance differences between different fabs?
Lip-Bu Tan:
That's a good question. In terms of Moore's Law, in the past we pursued doubling performance while also considering power consumption and cost. You can double performance, but cost and area cannot maintain the same advantage. So you have to make trade-offs in these aspects, unless you find new materials, new design methods, and actually implement them.
I've started recruiting more talent in materials science. This is the innovation focus in our field: how do we continue to advance?
I still remember 18 years ago, I was still investing in semiconductors. At that time, most VC firms, including some very good top-tier VCs, were my good friends. At the start of partner meetings, all the partners would be in the room listening to me talk about semiconductors. Halfway through, half of them would make excuses to leave. The remaining half would eventually ask: Lip-Bu, do you have any software-as-a-service projects? In the end, only two people stayed out of sympathy to listen to me.
So history has changed. Now semiconductors are hot again. Look at Jensen Huang's Nvidia, already a $5.3 trillion market cap company. Broadcom and TSMC are also at the $2 trillion market cap level. Lisa, my good friend at AMD, her company's market cap is close to $800 billion. And Intel is also approaching $600 billion.
So in a way, semiconductors are hot again and have become critically important. 15, 18, 20 years ago, when I invested in semiconductors, no VC wanted to invest with me, except for large companies like Samsung, Arm, SoftBank. Now I'm starting to see many VCs willing to invest in semiconductors, so I'm very happy.
Host:
Given the huge investor interest in this field now, whereas in the past this field was considered too difficult. You are both a long-term operator and have been doing venture capital at Walden for a long time. Generally, people have many concerns about semiconductor investment; let me list a few: it's very capital-intensive; tape-out success is highly unpredictable; you must deeply understand workloads; another factor is the high risk of customers switching suppliers.
We've been involved in some companies together that may have already secured design wins, but whether they can scale order volume is still a question. There's also cyclicality: you build heavy-asset manufacturing capacity, but demand might change in a given year, or it might not.
How do you view why this industry is difficult? At the same time, there is now long-term demand growth from different areas, like the recognition of the importance of supply chain diversification, and the explosive demand growth on the AI side. You are still an investor, and now you've made the biggest bet of your life by becoming CEO. How do you think about these different risks? How would you advise others to invest in this supply chain?
I know this is a very broad question, but considering your experience, I think many people might now have a "YOLO investing" mentality: for example, if there's a memory shortage, they buy memory stocks; but at the same time, they are unwilling to take on things that require a ten-year timeline, like materials science.
Lip-Bu Tan: Okay, your question covers a lot of ground. Let me try to explain.
First, venture capital and entrepreneurship are in my blood; I truly enjoy the process. This isn't to show off, but I do have some good exit cases. To date, I have 159 IPOs and 126 M&A deals, including in semiconductors. If you look only at semiconductors, I've invested in over 200 companies over the years, 38% of which are in the U.S. So I usually look at some micro trends.
Host:
Just to clarify, that's very impressive.
Lip-Bu Tan:
Thank you, thank you. I just enjoy the process of building these companies. But more importantly, on the investment side, what I look at first is: where is the bottleneck? What problem are you really trying to solve?
For example, I invested in a company called Credo Semiconductor, which has a lab in Australia. At the time, I saw that interconnect had become a bottleneck, so I decided to support it. I also supported Celestial AI, which does optical interconnects. Because within clusters, interconnect speed is becoming increasingly important, so I think optical technology will be very important. Look at Jensen Huang, he has invested in almost all photonics-related companies.
Also, I look at what solutions the market needs. For instance, we just talked about design complexity and cost—can we use AI and machine learning to drive better designs and better solutions? There are now several new startups entering EDA-related fields, trying to improve performance. I think this is a gold mine.
And new materials. We talked about indium phosphide, so I invested in Inphi, which was later acquired by Marvell. You can also invest in some new materials, like gallium nitride and silicon carbide. Some of these companies have already started being acquired, including a company doing power management called Empower, which does very well in IVR.
Power management has now become a bottleneck. For example, stepping down from 40 volts to 1 volt, a lot of power is lost in the conversion process; how to improve power efficiency is very critical. So power, heat dissipation—these have all become bottlenecks.
Therefore, I always start from "what problem are we really trying to solve?" Is this problem real? Are customers truly suffering because of it? If so, I start investing.
The next step is, from day one, lock in the first customer. I usually prefer the first customer to be a hyperscaler, because they have scale. If they like what you have, they are willing to pay millions of dollars over the next few years, or even give a purchase commitment. This is important, because with one large customer, you can expand.
So I always look at certain formulas: How do you achieve this? Where do you find talent? Sometimes, finding talent is very important. This is also why I am very interested in the US, Silicon Valley, and Austin. Additionally, Israel has a lot of talent. So I have invested in quite a few projects in Israel.
Because Israel has many disruptive, innovative entrepreneurs who work extremely hard. Even during wartime, they still hold conference calls. Sometimes they say: Okay, there's an alarm now, I have to go to the basement, the internet might not be good, maybe we can only use voice. To some extent, this is even a bit amusing. I deeply admire this resilient entrepreneurial spirit.
Overall, I think there are many opportunities, especially in the AI field. Now, besides Agentic AI, physical AI is also becoming the next huge frontier. You have to look at the problem from a full-stack perspective.
This is also why I am still deeply involved in many frontier models and some of the investment projects I support, because I am very bullish on open-source frontier technologies for physical AI. I think that is a gold mine.
Host:
You mentioned there is an opportunity to use AI to make certain links in chip design and testing faster, cheaper, and more creative. Combined with your experience at Cadence, where do you think the most fertile direction lies? Is there anything already starting to work?
Lip-Bu Tan:
I was at Cadence for about 15 years, and I am very happy. One thing I am very proud of is that I was able to find my successor along the way and mentor him. Later, he became a very outstanding CEO. He very much embraces AI, using Agentic AI to improve efficiency.
That is the good side. I think Synopsys is also working hard on these things. They received a $2 billion investment from NVIDIA, which I think can help them a lot. He also acquired Ansys to enter the entire system design field.
Overall, these companies are all trying their best. But startups also have opportunities to do something more disruptive, eventually either going public or being acquired by these two companies, or Siemens.
So I think opportunities belong to everyone, depending on the entrepreneur's vision. My philosophy has always been: if the entrepreneur wants to sell the company, because it's a faster exit path with no lock-up period and no need to worry about quarterly performance, that's fine too. There are also entrepreneurs who want to IPO from day one.
As VCs, I think the three of us are all VCs, we support the entrepreneurs' dreams and help them realize their dreams.
Host:
If you look at these different directions you mentioned, including future product development, or the impact of AI on the semiconductor industry, there are now companies like Periodic doing materials, and companies like Purepoint doing EDA and design, as well as other links in the manufacturing chain.
Do you think Intel ten years from now, or future semiconductor companies, will be fundamentally different from today because of AI? If so, where will the differences be?
Lip-Bu Tan:
I think so. First, going back to the issues you raised at the beginning: capital-intensive, unpredictable, cyclical. These must all be factored into your investment decisions.
I usually like to enter very early and build a team. This is interesting. I think you do the same. Second, you need to find the right investors to collaborate with you. Not always looking at brand-name institutions; I usually look more at the individual. Who is the person who truly understands this field? Most importantly, you need to find partners who can go through difficult times and good times together.
Many people are happy to work with you in good times, but as soon as the company runs into trouble, they leave. I like those who truly accompany the company through difficulties. Some successful companies nearly went bankrupt multiple times before finally taking off. So finding partners willing to do this is very important.
Additionally, look at strategic investors, whether they can help the company create value in manufacturing, memory, connectivity, etc. I also have some friends in the growth stage and hedge fund space; I like them very much because they have different perspectives. They understand the public market and can guide entrepreneurs on which paths not to take. These are all very helpful.
Overall, this is very interesting. You realize that entrepreneurship is actually just like engineering—it's all about problem-solving. At every step, you have to find people who can help you solve the problem. If solved, you move on to the next frontier.
Frankly, looking back, out of ten companies I've invested in, nine changed their business plan halfway through because the market changed. So I like entrepreneurs who are a team, not just a single person. Second, they must be open-minded, willing to listen, and willing to accept our coaching.
Finally, they form their own plan, rather than doing exactly what I say. The better state is that you give them enough feedback, and they draw their own conclusions. As long as you agree with their judgment, even if it differs from your thinking, it's acceptable. This is the interesting part of entrepreneurship. They can move forward faster.
Going back to your question, if we look ten years from now, what kind of company will win? This is just my personal opinion: companies that can clearly articulate their strategy, have laser focus on a specific niche, find the right partners, and possess the ability to scale will win.
To some extent, this goes back to my full-stack viewpoint. You need to have a full-stack solution. It could be a large company transforming into a big platform. Like Jensen Huang, I admire him very much. He focused on CUDA, focused on software libraries. He said, I want to be a platform company, and he indeed achieved it.
It could also be a startup, like Anthropic, OpenAI, which found a path in a more elegant way and changed the game. Startups move extremely fast, moving at the speed of light, and can also become dominant players.
I hope Intel can also play such a role, because we have XPU, NPU, advanced packaging, and foundry. If you put these together, you can build specialized chips for different workloads. I am moving in this direction.
Host:
That makes a lot of sense. Part of my question just now was to understand where you are heading, and the other part was to ask whether this will fundamentally change the way you work. Because in the software world, I see very big changes happening now: who you hire, who you want to join the team, many people are starting to manage multiple agents.
So now many people I know actually prefer to hire people in their 30s, 40s, and 50s, because they are used to managing teams. They think this can be directly transferred to managing agents, including understanding how to set up complex tasks, how to do QA, and so on.
I want to know, in the physical world, or in a fab environment, how do you view team structure, capability requirements, or the changes after AI is overlaid? Is this a naturally slow evolution, or will there be radical changes in certain areas? For example, in the materials field, is it now enough to just use these three models plus some chemistry knowledge? So I'm curious how you see that future world.
Lip-Bu Tan:
Good question. Going back to what I said earlier about "crawl, walk, run." In the "crawl" phase, you first need to recruit the best talent in the semiconductor industry. Now I am starting to think about what software talent I need to bring in to build the full stack.
Currently, the average age of my team is probably in the 40s and 50s; I need to bring in some new talent. They understand workloads, understand frontier models, understand open source—this is important.
Now my son has become my teacher. Every time he invites me to his house, while we play with the grandchildren, I ask him about AI and machine learning. He is deeper into it than I am, so I have learned a lot and am trying to understand investments and bring in some talent.
We are changing Intel. It used to be a very old-school, traditional, spreadsheet-dependent company. Now I am transforming it into an AI-enabled company, using AI in design and making the entire organization embrace AI. This way, it won't be so dependent on spreadsheets and manpower.
You must combine excellent talent with the best AI tools, not just for organizational management, not just for sales; now I am also starting to consider marketing, design, and other links, all embracing AI.
Host:
I think for many investors, at least for me, after founding a company in the past few years, thinking about different funding sources for capital-intensive companies has been a very educational process.
I have done a lot of software investing before. If you say, I need $150 million before reaching a certain key scale, then you need some very smart friends with completely different balance sheets.
You have experienced this for a long time. You also have a unique experience of working with the government as a large stakeholder. How do you view this kind of industrial policy? It once brought huge successes like TSMC, which is one of the most important companies in the world. But in American business culture, industrial policy has long been unpopular. How do you think this perception should change now? Where is it applicable?
Lip-Bu Tan:
That's a good question. Obviously, for capital-intensive businesses and infrastructure-type projects, you need access to capital. To some extent, for early-stage venture capital, many investments are now also becoming capital-intensive. In the past, a VC firm willing to invest $1 billion in a company was unheard of in the VC industry, but it is happening now.
So to some extent, you have to adapt. I like to look at it using a bell curve. Either you enter very early, because now a Series A might be valued at over $1 billion, so you have to enter at the pre-seed stage, before the company valuation reaches $2 billion or $3 billion. This is very rare today, so you have to pick the right projects.
The other part is finding capital that can help the company expand. This is also why some mutual funds are starting to be willing to enter the unlisted market and join me in investing in early-stage projects. I welcome them very much, because they are less sensitive to the requirement of "must hold 20% of the company." There isn't that much 20% to give away anymore. So you must find the right investors.
In capital-intensive areas, such as AI factories and foundry businesses, you do need to leverage government funds, sovereign funds, and some very large capital. Now there are some large funds specifically supporting infrastructure, and we also hope to utilize some of that capital to ensure we can expand operations.
Overall, governments and sovereign funds have become very important. At the same time, as a public company, I also intentionally pay attention to some longer-term, growth-oriented investors, because they can help me develop the business, rather than just focusing on short-term capital allocation, asking if you want to buy back shares. Those questions are also good, but at the same time, I still have to build the business. So balance is very important.
What Investors Misunderstand Most About Intel
Host:
What do you think investors misunderstand most about Intel at this point in time?
Lip-Bu Tan:
Quite a few things. First, going back to "crawl, walk, run." Over the past four months, I have been crawling. But people are already starting to realize its potential. Another very important point: we must truly deliver the best products. For example, in PC client, we still have market share. But we really need to build better performance. So I am quietly building CPU architecture, GPU architecture, and software architecture teams, allowing us to move faster like a culture composed of multiple startups, and achieve a leap with better technology.
Beyond products, there is some new energy pouring in, such as Agentic AI and Physical AI. These are huge markets we can invest in.
On the foundry side, we are still far behind TSMC, whether in terms of performance or other aspects. So we must remain humble and build those foundational blocks, like the IP, yield, defect density, cycle time I mentioned earlier, making it more efficient and reliable. Foundry is a trust business. Customers must trust you first before they will hand over their wafers and rely on you. So these things take longer.
But I think by 2030, 2031, people will start to see how much potential we have. On the product side, PC client is our base. Then we will enter edge computing, enter Physical AI and Agentic AI.
In the past, we mainly provided servers and PCs for humans. Now you will see another dimension: millions of agents, which also need computing power and need access to the software stack. So I think we have an opportunity to participate in this part.
The game is not over yet. We can still play our cards in Agentic AI and Physical AI. This is the direction I am heading.
AI has just begun. The training part is dominated by Jensen; edge computing, agents in Agentic AI, and Physical AI, I think, are all huge opportunities. Everyone has a chance. So this is the direction I want to pursue.
I hope investors will understand that although we have created a 6x return for shareholders over the past 14 months, this is just the beginning. We still have a lot of room.
Host:
There are still venture-capital-style returns from here.
Lip-Bu Tan:
Yes. I am always looking for 10x opportunities. As someone who is a venture capitalist at heart, you always want to find 10x.
At Cadence, when I was CEO, starting from the interim CEO starting point of $2.42, by the time I stepped down as Executive Chairman, I created roughly an 85x return for shareholders. Close to 76x, even 85x.
Achieving that at Intel is difficult because the base is larger. So I said, okay, let's look at 10x then. If we can achieve 10x in five or ten years, I think that's a very good return. As a VC at heart, that is my goal.
Will Compute Always Stay in Data Centers?
Host:
I wish you success in this very large mission on an already large base. There is an implicit judgment behind your description, which is where workloads will run. Some people say we will just build bigger and bigger data centers, 1 gigawatt is just the beginning. Centralized operation, even including centralized inference computing, will become the dominant approach in terms of efficiency.
But others also consider the edge and the client side. Do you believe there will be some equilibrium state for computing in the future? Or can it only be determined by the workload itself? What do you think?
Lip-Bu Tan:
That's a very good question. AI infrastructure is currently being built on a massive scale, and I think this is correct. I don't think it will slow down, because workloads are increasing significantly.
Host:
We are currently supply-constrained.
Lip-Bu Tan:
Yes, supply-constrained. So if anything will slow down development, it's the supply constraint.
But on the other hand, I always look at what solutions and applications all this infrastructure construction is ultimately meant to serve. I focus more on applications. If you can identify a massive application, or a few applications that together are meaningful enough, and concentrate around that, then not everyone involved in building will win. Some will win big, some will slowly fail, or just move sideways.
It’s just like the internet era. You could see some companies eventually became very large, like Amazon and Netflix; some became marginalized, disappeared, or got acquired. So for me, the thinking is the same. What you really need to focus on is: what application do they want to serve? How big is that application? Is it sustainable? Is it too crowded?
If it’s too crowded, maybe only one or two will remain in the end, and the rest will be consolidated. So this industry will go through massive growth, then start consolidating, and eventually maybe one or two companies become the real winners. We’ve seen this movie before, so I’m not surprised.
Focus on applications. Netflix is an application, Amazon is a real application. In my view, they are the winners.
Host:
But you’re assuming that some of these applications, served through clients or edge computing, would be better than relying entirely on data centers?
Lip-Bu Tan:
Exactly right.
Host:
I’ve invested in some robotics and defense companies myself, so I know on-device computing is a very important option. For example, if there’s a robot in the home in the future, what computing power and connectivity you assume at home will determine what you can do. I feel this was somewhat forgotten during the SaaS era.
Lip-Bu Tan:
Yes. My investment logic is: find the problem that truly needs to be solved. Second, find the players you can partner with. Third, look at the application. How big is this application? Is it sustainable? If it’s really big and you believe in it, then double down, triple down.
Host:
But that also includes betting on applications that haven’t been widely deployed yet.
Lip-Bu Tan:
Right.
Host:
Fantastic. Thank you so much for joining us today. It was a pleasure talking with you.
Lip-Bu Tan:
Thank you very much.
