Intel CEO Patrick Gelsinger stated that his return target for Intel is "a 10-fold increase within 5 to 10 years," and he is systematically rebuilding Intel’s technology roadmap around advanced packaging, novel semiconductor materials, and next-generation substrate technologies.
In a recent podcast episode, Chen Liwu detailed his roadmap for transforming Intel: after stabilizing the balance sheet and refocusing on product lines, he is now shifting emphasis toward advanced packaging technologies such as EMIB, glass substrates, and new materials including gallium nitride (GaN), silicon carbide (SiC), indium phosphide (InP), and synthetic diamond, to address the challenges posed by the physical limits of traditional process node scaling. He also revealed that the surge in agent AI and inference workloads is driving strong renewed demand for CPUs, with the CPU-to-GPU ratio in data center servers evolving from the previous 1:8 toward 1:4 or even lower.
Chen Liwu stated that in the past 14 months, approximately six times the return has been created for Intel shareholders, but "this is just the beginning." He expects that by 2030 to 2032, the market will begin to truly recognize Intel’s potential—not only in its traditional PC client base, but also extending into emerging markets such as edge computing, physical AI, and agent AI.
In his view, if Intel’s XPU, advanced packaging, and foundry capabilities are effectively integrated, they will enable customized chip solutions for diverse workloads—a long-term strategic direction he has set for the company.
New materials are the key to breaking through, with advanced packaging and glass substrates as the focus.
Amid the growing physical limitations of traditional process node scaling, Chen Liwu has turned his focus to materials science and advanced packaging. He stated that Intel has already begun mass production of the 18A process, is advancing toward mass production of 14A, and can see a technological roadmap extending to 10 nanometers and even 7 nanometers—but "this path will become increasingly expensive and increasingly difficult."
To this end, Chen Liwu has initiated several strategic moves in the field of packaging materials. He invested in the glass substrate company 3DGS, drawn to glass’s unique properties as a thermal dissipation and insulating material; in chip-to-chip interconnection, Intel is actively promoting its next-generation advanced packaging technology, EMIB, and has announced partnerships to advance advanced packaging manufacturing in India and New Mexico, USA. Intel holds approximately 1,000 patents in the module domain, and how to effectively integrate substrates with modules is the core engineering challenge emphasized by Chen Liwu.
In the area of new semiconductor materials, Chen Liwu stated that he has made investments in gallium nitride, silicon carbide, and indium phosphide, with some of these investments already acquired by major semiconductor companies such as ADI. He has also invested in a company specializing in synthetic diamond wafers, seeing strong potential for diamond as a thermal insulation material in chip packaging. "That's the spirit of an engineer—you keep hitting bottlenecks, then find ways to overcome or work around them," he said.
Contract manufacturing: Trust comes first; yield and cycle time are the key metrics.
Intel's foundry business was once viewed by outsiders as unsustainable, but Chen Liwu chose to stay the course. He stated that the core rationale behind this decision is that advanced manufacturing within the United States holds strategic value for supply chain security, and no major semiconductor company can afford to concentrate its supply chain in just one or two geographic regions.
At the operational level, Chen Liwu has set yield, defect density, and cycle time as the top priorities for the foundry business. He emphasized that foundry is fundamentally a business built on trust—"customers must trust you before they entrust you with their wafers." Once yield targets are not met, customers who suffer revenue losses will leave, and regaining them will be extremely difficult.
He also stated that Intel and TSMC are partners, not merely competitors, and that the industry as a whole needs more capacity to meet continuously growing demand. He expects Intel’s foundry business to begin realizing its true potential in the market by 2030 to 2032.
Terafab Partnership: Building Semiconductor Infrastructure with Musk
Chen Liwu revealed that the Terafab project, advanced by Intel and Elon Musk, stems from their shared assessment that the development of semiconductor infrastructure has lagged behind the growth in AI demand in terms of capacity, production efficiency, and power efficiency. Under this collaboration framework, Musk decided to build his own wafer fab, with Intel providing technical and process support to help accelerate production. Chen Liwu said he holds weekly meetings with Musk’s team, and the collaboration is progressing smoothly.
He also mentioned that Musk has an unconventional approach to operations, such as once discussing whether smoking should be allowed in certain areas of the clean room: "I might not go that far, but perhaps certain areas could be permitted—the key is to maintain an open mindset."
Investors' biggest misconception: Intel is still in its "crawling" phase; its true potential will emerge after 2030.
In response to market skepticism about Intel’s transformation progress, Chen Liwu invoked his longstanding "crawl-walk-run" framework. He stated that the past several months have still been in the "crawl" phase: Intel is quietly building teams for CPU architecture, GPU architecture, and software architecture, aiming to drive breakthrough innovation at the pace of a large startup; on the foundry side, the gap with TSMC remains significant, and Intel must remain humble while strengthening foundational capabilities such as IP and yield.
“My VC intuition tells me to look for 10x return opportunities,” said Chen Liwu. He referenced his experience at Cadence, where he generated approximately 76 to 85 times the return for shareholders from his time as interim CEO until his departure. He acknowledged that Intel is larger and harder to replicate, but “achieving a 10x return within 5 to 10 years” is his clearly defined goal.
Below is the transcript of the interview:
Host: Welcome back to No Priors. Today, Allad and I are joined by Liwu Chen—a legendary investor from Walden, former CEO of Cadence, and current CEO of Intel. We’ll discuss his plan to transform Intel, what it means for the U.S. government to become a major shareholder, how to become an outstanding semiconductor investor, and whether we can manufacture chips within the United States. Welcome, Liwu.
Why take on the burden of Intel?
Moderator: We’ll start with an obvious question. Leading this critically important U.S. semiconductor company is an incredibly challenging role. Why did you take it?
Chen Liwu: That’s a great question. I’m 66 this year, and many people say I should retire and not take on the hottest job in the industry. There are a few reasons: first, this is a landmark company that is critically important to the entire semiconductor ecosystem and to the United States; second, after Cadence, I decided to take on one more major challenge.
Host: A lot has happened over the past year. What surprised you the most?
Chen Liwu: The most unexpected thing was something I had never experienced in any previous job or training—early one morning, President Trump asked me to resign, citing a conflict of interest with no exceptions. At first, I convinced myself: I don’t need this job; I’m doing this purely to save Intel. After setting aside my personal emotions, I began to think about what I could do for Intel. Fortunately, I secured a meeting on Thursday morning and another on Monday, during which I shared my story with him—I was born in Malaysia, raised in Singapore, graduated from MIT, and have lived in the U.S. ever since without ever leaving. I shared these details with him, and he listened, ultimately giving me the chance to continue. I am deeply grateful.
Host: You said this job is about "saving Intel." What does a winning, thriving Intel look like in your mind?
Chen Liwu: It’s been 14 months, and a lot has happened. First, I changed the culture, clarified accountability, and accelerated decision-making. I was used to the pace of startups, where everything moves at light speed, but Intel had layer upon layer of bureaucratic meetings—I had to change that. Second, I listened to customers—truly satisfying customers requires humility, a willingness to listen, and the determination to confront and solve their problems. Third, from day one, I decided that all engineering teams would report directly to me. As an engineer myself, I needed to know firsthand what was going wrong and what needed fixing. Listening to customers, satisfying them, ensuring we have the right products, streamlining our product line, and establishing a clear roadmap and vision for the next five to ten years.
Intel's Ten-Year Vision
Host: What is your vision for Intel ten years from now?
Chen Liwu: My consistent approach—whether at Cadence or Intel—is to start by crawling, stay humble, and listen to customers; then walk; and finally run. One step at a time.
The first step was to strengthen the balance sheet—truthfully, the balance sheet was in very poor shape at the time. I was relieved that the U.S. government became the majority shareholder. I explained to President Trump: Look at Japan, look at Singapore—this is infrastructure-level support, and the government should provide it.
Second, I am deeply grateful to my old friend Jensen Huang—he invested $5 billion in Intel—and I’m pleased that my efforts have helped grow that $5 billion into $25 billion or more. Additionally, Masayoshi Son of SoftBank, where I previously served on the board, also came to our aid. Through these contributions, we strengthened our balance sheet.
Next is focusing on products, streamlining the product line, listening to customers, and launching the next generation of leading products. Fortunately, demand for agent AI and inference CPUs is currently extremely high, so in a sense, I’ve caught the right moment. Previously, the ratio of CPUs to GPUs during training was about 1:8; now I’m seeing it shift to 1:4 or even lower. CPUs have become important, and I’m pleased to see this.
I’ve spoken with several AI model developers who said that, in the reinforcement learning phase and in coordinating the speed of all agents, CPUs actually perform better. As a result, my demand for CPUs is now very high. After establishing a solid foundation in our data center server product line, another key business is our foundry operations. This is a capital-intensive business and not easy. You need the right IP portfolio—for example, low-power IP for mobile customers—without which you can’t serve them. It’s a service business and a trust-based business—if yield isn’t up to standard, customers will abandon you due to revenue losses. That’s why I’m intensely focused on yield, defect density, and cycle time to ensure we serve customers with high quality and reliability. Ultimately, we must move toward a full-stack approach—not just silicon itself—you need software too. Some customers directly ask me, “Give me an entire rack,” and you must deliver system-level solutions. I’m quietly advancing on all these fronts while recruiting the best talent I can find. By the way, all hiring is done personally by me—I don’t use recruitment agencies.
Partnering with Elon Musk at Terafab
Host: Another significant initiative that has generated widespread discussion is Terafab and its collaboration with Elon Musk. Could you explain how this partnership came about and how you work together?
Chen Liwu: I believe we all agree that Elon Musk is one of the greatest entrepreneurs of this century. He and I share a common assessment: semiconductor infrastructure has not kept pace with the growth of AI—there are gaps in capacity, production efficiency, and power efficiency, and we both recognize this issue.
Second, I truly enjoyed collaborating with him. He is highly unconventional and consistently questions, “Why do things the traditional way?”—which was refreshing. I appreciate hearing different perspectives and working together to find the optimal path, as both sides learn a great deal. He has a clear vision: his robots and his car require vast amounts of chips.
Specifically, Terafab decided to build his own wafer fab, and we’re very happy to collaborate with him to help accelerate production by leveraging some of our technologies and processes—it’s a joint initiative. His team is excellent; I meet with them weekly, and working with him is incredibly exciting. He’s mentioned some unconventional ideas, like allowing smoking inside cleanrooms—I might not go that far, but perhaps certain areas could accommodate it. The key is maintaining an open mindset, and we’re seriously listening and evaluating these suggestions.
Changes in the global semiconductor supply chain
Moderator: From a macro perspective, how is AI driving changes in the global semiconductor supply chain? What observations do you have when looking country by country?
Chen Liwu: The impact of AI on the entire landscape will surpass that of the internet and be even more profound. AI first enables you to accomplish tasks more efficiently; with the assistance of numerous intelligent agents, many previously tedious tasks that required manual effort can now be completed much faster. For example, in the field of semiconductor design, timing optimization and time-to-market can be significantly accelerated, while costs are also reduced.
Demand for AI faces several bottlenecks: first, power constraints—some countries simply lack sufficient electricity; second, the impact of helium, which many fail to recognize as significantly affecting the semiconductor industry; and third, memory shortages, the most pressing issue today—even if you ramp up production now, new capacity will take years to come online; CPUs and GPUs are also in short supply, driving prices up, with costs ultimately passed on to end users.
The companies most affected are those that do not embrace AI. AI can enhance efficiency across nearly all functional areas of a business; companies should proactively adopt AI and find better ways to leverage it—for forecasting, design, and various workloads.
Host: The simplest argument against Terafab and Intel Foundry’s competitiveness is the issue of labor costs and the feasibility of domestic manufacturing. What is the logic behind your decision to continue investing in the foundry business?
Chen Liwu: When I was deciding whether to continue investing in contract manufacturing or exit it, there was a lot of external noise—many said it was too expensive, that it wouldn’t work. But in the end, I concluded that this is extremely important for the United States and for the entire industry.
We have all experienced supply chain challenges; any major semiconductor company must seriously consider its supply chain and ensure it is robust and resilient, avoiding complete reliance on one or two geographically concentrated suppliers. An increasing number of people are recognizing that manufacturing in the United States is crucial.
Our most advanced processes, such as 18A, which is at the 1.4-nanometer level, are already being planned for 1-nanometer and 0.7-nanometer nodes. As process nodes shrink further, line widths become finer than a human hair, and complexity reaches unprecedented levels—any single error can render all prior work useless. For this reason, manufacturing precision requirements are continuously increasing, making it an ever-growing bottleneck.
We deeply respect TSMC; we are excellent partners, and the industry needs more capacity to serve customers, so we decided to push through—this is a crucial move in the long term and where I can create the most value for the industry.
Physical Limits and Advanced Packaging
Host: People have been discussing for a long time that chip scaling will hit physical limits—when linewidths become too narrow to shrink further. When do you think we’ll truly hit that wall?
Chen Liwu: We currently have 18A in production and are advancing toward mass production of 14A. I can see a path forward for 10 nanometers and 7 nanometers—this path is achievable, but it will become increasingly expensive and challenging. That’s why we need partners and must work closely with substrate suppliers and equipment manufacturers to jointly improve yield and performance.
Another critical area becoming a bottleneck is advanced packaging. TSMC has CoWoS, and we have a next-generation solution called EMIB; I must ensure it achieves the yield levels required by customers at mass production.
When traditional scaling began to hit its limits, I turned back to the materials level to seek breakthroughs—I invested in gallium nitride, silicon carbide, and indium phosphide. In terms of packaging materials, I began focusing on glass—glass is an excellent thermal insulator—and I invested in a company called 3DGS. Intel holds around 1,000 patents related to modules; integrating substrates and modules remains a critical challenge. We have also recently announced advanced packaging manufacturing collaborations in India and New Mexico, USA. Additionally, I am exploring synthetic diamonds—another outstanding thermal insulator—and I have invested in a diamond wafer company.
That’s the engineer’s mindset—you keep hitting roadblocks and then find ways to overcome or work around them. I’m now thrilled to apply the experience I’ve gained from deeply participating in every stage of the semiconductor lifecycle, from EDA tools to design to manufacturing, to make meaningful contributions to the industry.
Host: Is there a possibility that the convergence of process nodes could narrow the performance gap between different foundries, forming some kind of asymptote?
Chen Liwu: The essence of Moore’s Law is the doubling of transistor density, but power consumption and cost do not necessarily decrease proportionally year over year—you can double performance, but area and cost may not decline equally. Unless you discover new materials or new design methodologies. This is precisely why I’ve increased our hiring of materials science talent—this has become the core of innovation in this field.
Eighteen years ago, when I was investing in semiconductors, many top-tier VCs had zero interest in the field. I still remember that after presenting on semiconductors at a partnership meeting, half the partners found excuses to leave, and the remaining half asked, “Do you have any software or services deals?” In the end, only one or two people stayed behind out of sympathy. Today, NVIDIA, led by Jensen Huang, has a market cap of $5.3 trillion; Broadcom and TSMC are each worth $2 trillion; my good friend Lisa Su’s AMD is nearing $800 billion, and Intel is close to $600 billion. Semiconductors have once again become a hot field—an essential foundation. Fifteen to twenty years ago, almost no VCs were willing to invest in semiconductors with me, except for large institutions like Samsung, ARM, and SoftBank. Now, VCs are flooding in with tremendous enthusiasm for this sector, and I’m truly delighted.
Challenges in Semiconductor Investment
Host: You are both a long-term investor and an operator. Semiconductor investment faces many challenges—capital-intensive, unpredictable outcomes, the need for deep understanding of workloads, high switching costs for customers, and strong industry cyclicality... How do you view these risks, and how would you advise others on where to invest within this supply chain?
Chen Liwu: Venture entrepreneurship is in my blood—I truly enjoy it. I’m not here to boast, but for context: I have a track record of 159 IPOs and 126 merger and acquisition exits, with over 200 investments in semiconductors, 38% of which were in the United States.
In my investment approach, I always start with one core question: Where is the bottleneck, and what problem are you solving? For example, I invested in Cradle Semiconductor because interconnects had become the bottleneck; I invested in Celestial AI because optical interconnects are becoming increasingly critical within clusters—this is no coincidence that Jensen Huang has invested in nearly all photonics-related companies.
At the design level, can AI and machine learning help reduce complexity and improve design quality—I believe there is tremendous opportunity in the EDA space, with several startups actively pursuing this direction, making it a gold mine. In terms of new materials, gallium nitride, silicon carbide, and indium phosphide are all areas I’m investing in, with some companies already acquired by large firms like ADI. Power management—specifically the highly inefficient process of stepping down from 40V to 1V—is another bottleneck sector I see great potential in.
My investment framework has always been: Is the problem real? Are customers genuinely struggling with it? And then, critically: Who is the first target customer? I tend to favor hyper-scale customers—they have the capacity and willingness to spend millions over the next few years, or even offer some form of guarantee, if they like your solution, because landing one large client enables scalability.
Talent is also crucial—I am closely focusing on the United States, Silicon Valley, Austin, and Israel. Israel is home to highly disruptive, innovative entrepreneurs who work extremely hard. Even during wartime, they continue holding meetings—sometimes saying, “There’s an alert; I’m going to the basement—the internet might be down, so let’s switch to voice call.” This resilience deeply impresses me.
Beyond agent AI, physical AI is the next major frontier—you must carefully examine the full stack. That’s why I remain deeply involved in investments related to many cutting-edge models—I have strong confidence in open-source frontier technologies for physical AI; this is a goldmine.
Cadence's experience
Host: You mentioned that AI offers the potential for faster, cheaper, and more creative approaches to chip design and testing. Based on your experience at Cadence, which areas do you see as the most promising? Are there any applications already delivering results?
Chen Liwu: I spent nearly 15 years at Cadence, and one of my proudest achievements was identifying and personally mentoring my successor, who is now an outstanding CEO actively embracing AI by integrating agent-based AI into tools to enhance efficiency. Similarly, Sassine at Synopsys is doing the same, backed by a $2 billion investment from NVIDIA and an acquisition of Ansys to expand into full-system design.
Large companies are doing it, but there are still opportunities for startups to pursue more disruptive initiatives, which could eventually lead to an IPO or acquisition by one of the two giants. It depends on the entrepreneur’s vision. My consistent philosophy is: if entrepreneurs want a quick exit, we help them achieve that; if they aim for an IPO from day one, we support them on that path. As VCs, we back entrepreneurs’ dreams and help them realize them.
Scaling and Investment Decisions
Host: You mentioned these areas—materials companies, EDA, manufacturing. Looking ahead 10 years, will Intel or future semiconductor companies be unrecognizable due to AI?
Chen Liwu: I believe it will. Returning to the characteristics you mentioned—capital-intensive, unpredictable, and cyclical—these must all be factored into investment decisions. I typically prefer to enter early and build a strong team; find the right investors who will stand by you through tough times, not just fair-weather friends; and seek strategic investors who can add value to the company in areas such as manufacturing, storage, interconnectivity, or other dimensions. I also have friends in growth-stage and hedge funds who bring unique perspectives on public markets and can help entrepreneurs identify which directions to avoid—this is extremely valuable.
To be honest, when I look back, nine out of the ten companies I invested in changed their business plans along the way because the market shifted. That’s why I prefer entrepreneurs with a team rather than someone working alone. They also need an open mindset—willing to listen and consider our advice, but ultimately make their own judgments. The best outcome isn’t “I’ll just do whatever you tell me to do”; it’s when you provide enough feedback, and they independently arrive at a conclusion you recognize or understand—that’s the real joy of entrepreneurship.
Looking back ten years from now, the winners will be companies that focus on a niche area, find the right partners, and can scale effectively. Having a full-stack solution is crucial. Large companies can follow NVIDIA’s lead—like Jensen Huang did with CUDA and the platform—and fully commit to becoming a platform company; he succeeded. Startups can also revolutionize the game in more elegant ways, like Anthropic and OpenAI have done; startups can move at lightning speed and truly become dominant players.
For Intel, I envision it playing this role—we have XPU, advanced packaging, and foundry services; if we integrate these to create customized chips for different workloads, that’s my direction.
Team Restructuring in the Age of AI
Moderator: The software industry is undergoing significant changes—particularly in terms of who to hire and who is best suited to manage multiple agents. Many now prefer hiring individuals aged 30 to 50, as they are accustomed to managing teams, a skill that can be directly applied to managing agents. In the context of hardware or contract manufacturing, how do you view changes in team structure and capabilities?
Chen Liwu: Returning to the crawl-walk-run framework. During the "crawl" phase, I recruited the best talent from the semiconductor industry; now I’m beginning to think about what kind of software talent I need to bring in to build full-stack capabilities. At the same time, I’ve noticed that the average age of my team is in their late 40s to 50s, so I need to bring in some younger talent who can understand workloads and stay current with cutting-edge open-source models.
Interestingly, my son has now become my teacher. Every time I visit his place to play with my grandson, I ask him questions about AI and machine learning—he understands these topics better than I do. I’ve learned a lot and try to apply that knowledge to investment decisions and talent acquisition.
Intel was once a very traditional company heavily reliant on spreadsheets, and I am transforming it into an AI-powered enterprise—embracing AI across the entire organization, not just in design, to reduce dependence on spreadsheets. We are combining seasoned technical talent with AI tools, actively adopting AI not only in sales and marketing but now also in design.
Industrial policy and capital sources
Host: For capital-intensive companies, securing funding has always been a major challenge. Industrial policies have given rise to critical companies like TSMC, but this approach has long been unpopular in American business culture. What are your thoughts on this?
Chen Liwu: For capital-intensive businesses and infrastructure projects, access to capital is critical. Now, some VCs are willing to invest $1 billion in a single company, which was unimaginable before. Therefore, in early-stage investment strategies, you either enter extremely early, when valuations are still reasonable, or wait until Series A—but now Series A valuations have exceeded $1 billion, making it very difficult.
I warmly welcome institutional capital, such as mutual funds, whose sensitivity to ownership stakes is lower. For capital-intensive projects like AI factories or contract manufacturing facilities, support from government funding, sovereign wealth funds, or large infrastructure funds is essential. Sovereign wealth funds and government capital will become increasingly important.
As a publicly traded company, I consciously focus on investors with a long-term growth orientation rather than short-term capital that constantly asks, “When will you repurchase shares?”—of course, shareholder returns are a legitimate concern, but I must also build the business; finding this balance is crucial.
The biggest misconception investors have about Intel
Host: What do you think is the biggest misconception investors have about Intel right now?
Chen Liwu: There are a few points. First, returning to crawl-walk-run: Over the past few months, I’ve still been crawling, but people are beginning to see the potential. In terms of products, we still hold market share in the PC client, but we must significantly enhance performance—so I’m quietly building teams for CPU architecture, GPU architecture, and software architecture, preparing for a leap forward, moving as quickly as a large startup and leveraging superior technology to achieve that leap.
In manufacturing, we still have a significant gap with TSMC. We must remain humble and focus on building a strong foundation—IP, yield, defect density, cycle time—to make manufacturing more efficient and reliable. This is a business built on trust; customers must trust you before they entrust you with their wafers. These things take time, but I believe by 2030 to 2032, people will begin to see just how great Intel’s true potential is.
The PC client is our foundation, but we are extending to the edges, toward physical AI and agent AI. In the past, you only provided servers and PCs to humans, but now there’s a whole new dimension—millions of agents need access to compute power and software stacks. I believe Intel has opportunities in both agent AI and physical AI—this game isn’t over yet.
AI is still just the beginning—you have training powered by Jensen Huang, edge devices, intelligent agents, and physical AI—this is a massive opportunity, and everyone still has a chance. That’s the direction I’m fully committing to. Over the past 14 months, we’ve delivered six times the return for shareholders, but this is only the beginning—there’s still tremendous room to grow.
My VC instincts tell me to seek 10x return opportunities. At Cadence, I went from acting CEO to retirement, with the stock rising from $2.40 to deliver approximately 76x returns for shareholders; by the end of my term as Executive Chairman, it was around 85x. Intel is larger and harder to replicate, but my target is 10x—achieving a 10x return within five to ten years. As someone at heart who is a VC, that’s my goal.
Where will hashing power be located?
Host: There’s a view that data centers will keep growing larger, with gigawatts just the beginning, and centralization will dominate. But the business landscape you described also includes edge and client-side computing. How do you think computing power will ultimately be distributed among data centers, edge, and client devices—or will it be entirely determined by application workloads?
Chen Liwu: The current large-scale AI infrastructure development is correct; I see no reason for slowdown, as workloads continue to grow. The current constraint is primarily on the supply side—any slowdown stems from supply-side limitations, not demand-side factors.
But what I’m more focused on is: what kinds of applications will run on top of all this infrastructure once it’s built? You need to find applications with real scale—just as, during the internet era, companies like Amazon and Netflix rose to prominence while others vanished or were acquired. The AI industry will go through the same process: massive growth followed by consolidation, ultimately leaving one or two true winners standing.
Focusing on applications is key—Netflix is a true application, Amazon is a true application, and both succeeded. Moreover, certain applications are better suited to run on the edge or at the client side—such as robotics and defense—where local computing power is critical. Your assumptions about connectivity and the built-in capabilities of devices determine what you can achieve. This point was somewhat overlooked during the SaaS era.
My investment approach has always been: identify real problems, find the right partners, and assess whether the market size for the application is sustainable—if you truly believe in it, double or triple down. Of course, this also includes betting on applications that haven’t yet been widely adopted.
Host: Thank you so much for joining us today—it’s truly been a pleasure.
Chen Liwu: Thank you for the invitation.
Source: Wall Street Journal