Written by: Techub News Compilation

Introduction

Recently, NVIDIA founder and CEO Jensen Huang had an in-depth conversation with former U.S. Secretary of State Condoleezza Rice during the "Only In America" series at the Hoover Institution at Stanford University. The interview took place at NVIDIA's new headquarters in Silicon Valley, designed through supercomputer simulations. Huang not only shared his personal journey from immigrating from Taiwan to the U.S., attending a boarding school in Kentucky, to founding one of the world's most influential tech companies, but also for the first time in a public setting, retraced in detail the core idea behind the establishment of NVIDIA — the birth of "accelerated computing" — and how this vision has endured for thirty years, evolving from gaming graphics to today’s artificial intelligence revolution. This dialogue is not just a retrospective of an outstanding entrepreneur's journey but also a profound discussion on technological innovation, risk-taking, and the American spirit.

Summary

• Huang reflected on his early experiences as an immigrant, believing that America's freedom, rule of law, and opportunities are the foundation of his success.
• NVIDIA was founded based on a "first principles" judgment: general CPUs cannot solve all computing problems, especially the parallel computing tasks of simulating the world.
• The company chose 3D gaming as the first application for GPUs to solve the "chicken-and-egg" ecological dilemma faced by the new architecture.
• The rise of AI is not accidental; its underlying computational demands (simulating intelligence) naturally align with the parallel computing architecture of GPUs, representing an inevitable extension of a long-term technological trajectory.
• Huang holds a "cautiously optimistic" view on AI and believes that the key to national competition lies in innovation at the top "application layer."

From Taiwan to Kentucky: The Starting Point of an Immigrant Story

Huang's story begins with a journey across Asia. He was born in Taiwan and moved with his family to Thailand at the age of five due to his father's job change. In 1973, after a coup in Thailand, the situation became unstable, and his parents decided to send 9-year-old Huang and his 10-year-old brother to the U.S. to live with their uncle in Washington State for safety. His first impression of America was filled with novelty: "I had never seen a house with carpet; it felt like walking on a bed with shoes on." From breakfast cereal, to the TV shows "Speed Racer" and "The Partridge Family," to Snickers bars, everything felt like a miracle to him as a child.

After a short three-month stay at his uncle's house, due to financial constraints, his parents sent them to a low-cost boarding school in Kentucky. Huang recalled that it was a small town called Oneida with a population of only about 600 people. "If you Google Map Oneida, the beauty of it is that there is nothing around it; it's just a little dot." Upon arriving, as the only Asian face in town, he inevitably encountered prejudice and curious stares. However, the nature of children quickly helped him integrate into the group; he joined the swim team and soccer team and fell in love with American foods like sausage gravy and hamburgers. After a swimming meet, the restaurant they went to — "like a spaceship" — McDonald's became the "best restaurant in the world" in his mind.

This experience shaped Huang's understanding of "expectation." He believes that immigrants often arrive with great hopes and dreams but have low expectations regarding specific aspects of life, leading to a deep gratitude for everything they receive. "When you come from more difficult circumstances, you appreciate everything. You are here by choice; you want to be here." His father had a chance to train in New York in his early years and hoped that the family could eventually come to this "incredible land" of America.

Education, Encounters, and the Silicon Valley Enlightenment

After moving to Oregon, Huang's growth trajectory resembled many teenagers who loved math and science. He jokingly remarked that in high school, those who liked math and science usually made only two or three other like-minded friends. They participated in the math club, science club, and of course, the computer club. In their spare time, they would hang out at arcade games and pinball machines.

Regarding college, his initial plans were quite practical: to follow in the footsteps of his family (his parents and brother had attended the same school) and enroll in Oregon State University for engineering. The university had a good engineering program, and the costs were affordable. It was here that he met his future wife, Lori. At the time, out of 250 students in the class, only 3 were girls. The 16-year-old Huang applied a bit of "strategy" by managing to get into the lab class that Lori attended, reducing the competition from 250 to 4. His ultimate pickup line was, "Do you want to see my homework?" This marked the beginning of their relationship, which continues to this day.

The call of Silicon Valley came from campus recruitment. AMD's recruiters came to Oregon State University, offered Huang a job, and promised to support his simultaneous studies at Stanford University. "Wait a minute," he recalled his reaction, "I can work at AMD, you’ll pay me a good salary, and you’re also willing to pay for me to attend Stanford?" After receiving a positive response, he eagerly accepted this "dream come true" opportunity. A year later, Lori graduated, and they got married, embarking on their life together. Huang worked and studied at Stanford, a process that took about eight years intermittently. He joked that he was "the student who spent the longest time paying at Stanford."

This experience of studying while working profoundly impacted him. "When you’re in school, you feel that academics are very theoretical because you’re not sure if learning these things has any purpose or benefit. But the benefits of working while studying, especially at Stanford, allowed me to see how important the principles taught were in everything I do today." Stanford shaped his view of computer science and its impact on the industry, particularly the intersection of technology, fundamental science, and computer science strategy, laying the ideological foundation for his later founding of NVIDIA.

Founding NVIDIA: An "Impossible" Business Plan

In the early 1990s, as the PC revolution was just beginning and Moore's Law was driving the era of CPUs, Silicon Valley was focused on general computing. Huang and co-founder Chris Malachowsky, however, saw a different future. They proposed two core ideas based on "first principles."

The first was a technological vision: they believed that many significant problems, such as real-time computer graphics (one of the toughest problems in computer science at the time) and simulations, could not be effectively handled by general CPUs. They envisioned that there should be a way to "accelerate" the CPU, offloading tasks that were not suitable for general computing. "It's like in your house, the kitchen has only one tool, and the garage has only one tool. The company has only one tool. Every job requires the right tool." They believed there existed another tool that could enhance the CPU, making the computer essentially a supercomputer.

The second was a commercial and ecological challenge: how to get developers to accept this new architecture? The CPU ecosystem had been developed over decades, building a vast application base and forming a strong positive feedback loop (increased installed base -> rising sales -> more applications -> reinforced cycle). How to break this "chicken-and-egg" dilemma? Historically, apart from NVIDIA, few other architectures have succeeded in challenging the CPU’s dominance.

Therefore, the key question was: What was the first "killer application"? They ultimately set their sights on 3D graphics. However, the professional computer graphics market at the time was minuscule, with almost only Silicon Graphics Inc. (SGI) in existence. So they thought of a more scalable potential field: video games. "Computer games are not only fun but might push the boundaries of computing power more than any other application," Huang pointed out, noting that the upgrade drivers of most consumer PCs came not from adults but from children wanting more powerful gaming machines. The gaming market had a large capacity and an extreme demand for graphics performance, enough to support the initial development and iteration of a new architecture.

However, this business plan still sounded "impossible": it needed to resolve multiple "chicken-and-egg" issues, required substantial funding, and attract world-class computer scientists. But Silicon Valley's venture capital culture made it possible. "I still remember explaining this plan to everyone... yet, Silicon Valley is here, and Sand Hill Road will fund me." Institutions like Sutter Hill Ventures became early investors in NVIDIA. Huang summarized, "We just determined based on first principles that general computers could not become the only computing platform; if we introduced this new idea, it could solve so many interesting problems."

From Graphics to AI: The Victory of Long-Termism

NVIDIA’s GPUs emphasized programmability from the start, laying the groundwork for its later generalized applications. Huang recalled that the GeForce 3 graphics processor had 57 million transistors, more than the combined total of Pentium 4 and Pentium 3 at the time, and for the first time, featured a CPU-like instruction set for game programmers to create effects.

For the next thirty years, NVIDIA has been "swimming upstream." The turning point was that the "simulated world" problem they solved resonated with artificial intelligence, especially deep learning, in its underlying computing patterns. "Computer graphics is essentially a simulation of the world. And in many ways, artificial intelligence is a simulation of thought and the brain. Simulated computations can (though not entirely but largely) be performed in parallel." The processor architecture used for simulations is different from that used for stepwise execution tasks (like executing a recipe). The world happens simultaneously and in parallel.

Based on this principle, NVIDIA continually sought new problems suited for GPU acceleration: after graphics came seismic processing (inverse physics), CT reconstruction, ultrasound, molecular dynamics (Newtonian physics), and so on. Then one day, deep learning researchers (like Stanford's Andrew Ng, the University of Toronto's Geoff Hinton, and NYU's Yann LeCun) found them. Huang keenly realized this was a field where they could make significant contributions. The collaborative results were remarkable, achieving breakthroughs in computer vision and other areas with GPU-accelerated deep learning.

Success triggered deeper contemplation: why was it effective? What else could be done? How far could it go? What does it mean for computer science and the entire industry? NVIDIA gradually deconstructed everything to first principles, then reconstructed the company's position in the future world. "All of this is about reasoning, vision, strategy, discipline, patience," Huang said, "and belief." He acknowledged that the company had "come a long way through hardships." "Every step we took was very difficult. The reason we’re here today is that no one believed in it. We were fortunate to build for a full decade without anyone paying attention."

During those times without external positive feedback, how did he maintain team morale? Huang's answer was to return to core values and beliefs. "You must believe in what you're doing. You have to go back to your core foundation... You must see that future in your mind, even if others can't. You have to tell that story so others can see it in their minds, too. And you must believe in it yourself."

Cautious Optimism in the AI Era and National Competition

As an infrastructure provider for the AI revolution, Huang holds a "cautiously optimistic" view towards this technology. He believes that intelligence is a fundamental element of society, industry, and all human endeavors that must be accelerated. But at the same time, caution is needed to ensure that AI operates as promised and functions reliably, avoiding the creation of systems that appear intelligent but are fundamentally flawed. "Reliable functioning is safer. I want my car to operate as promised. AI also needs to operate as promised."

When discussing the international competitive landscape of AI, he likened the tech stack to a "five-layer cake," with each layer being crucial:

1. Energy Layer: Electricity and land.
2. Chip Layer: The field NVIDIA is in.
3. Infrastructure Layer: Cloud services.
4. AI Model Layer: Currently the focus of public discourse.
5. Application Layer: Specific areas where AI is applied, such as healthcare, military, national defense, cybersecurity, transportation, and manufacturing.

Huang emphasized that while the model layer attracts much attention, the most crucial aspect for the nation is actually the top application layer. "It is this layer that will drive our industries forward." He cautioned that while the U.S. is currently in a leading position, changes in leadership could occur during pivotal technological shifts. Policies must ensure that the most vital innovations in the application layer are not hindered, as the country that achieves the greatest advancements in this layer will maximize the industrial revolution's benefits.

The "American Dream" and the Soil of "Only in America"

Reflecting on his personal journey and that of the company, Huang largely attributes success to America's unique environment. When asked if all of this could only happen "only in America," he described it as a series of "events with an extremely low probability" strung together. And what the U.S. provides is not headwind, but tailwind.

This "tailwind" includes: understandable and reliable laws and rules; a business and industrial environment where people adhere to rules, allowing you to find unmet market needs; the ability to trust that you can create great things and push them to market in a predictable way without facing arbitrary, capricious, or unknowable blockades. "These elements that entrepreneurs rely on thrive here."

Huang draws a parallel between entrepreneurial spirit and the spirit of immigration: "As an immigrant, you choose to come here. Compared to where you came from, you witness a miracle... the resources are incredibly abundant. You work hard because you yearn for success. As an entrepreneur, you also yearn for success. If you don't strive every day, you will perish. The spirit of entrepreneurship and the spirit of immigration are very similar." He admitted that his relentless "sense of despair" to make NVIDIA continually better echoes his parents' feelings of providing for the family. "Beyond hard work, there is no reliance, no way out."

"I can't imagine another place where all of this is possible," Huang concluded, "NVIDIA in many ways is a 'only in America' story. I myself am also a 'only in America' story. This all happens within a lifetime, not a fifth generation, nor a third generation, but within one person’s experience, across generations." His parents gave up everything to come to America, with no way back, sacrificing their lives to create more opportunities for their children; while the nation created opportunities, resources, systems, frameworks, and foundations for companies like NVIDIA to emerge. "I am the embodiment of the American Dream."

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。