By Carlos Taylhardat | November 5, 2025

In 1993, three engineers met over coffee at a Denny’s in East San Jose. They sketched plans on napkins, imagining chips that would do more than render graphics. A few years later, in Taiwan, a quiet industrial park became the home of a company that would make those chips real. By 2025, one of the companies was valued at more than $4 trillion, and the other was the hidden engine behind it. This is the story they rarely tell.

Origins: Two Companies, One Hardware Story

NVIDIA — From a diner to the GPU throne

NVIDIA was founded on April 5 1993 by Jensen Huang, Chris Malachowsky and Curtis Priem. Wikipedia+2Wikipedia+2
The three met frequently in late 1992 at a Denny’s roadside diner in East San Jose, California, to sketch out their vision of building graphics‑chips for the next wave of computing. Wikipedia+1
At the time, Huang was 30 and had recently earned a Master’s from Stanford and been working in chip design; Malachowsky and Priem brought engineering experience from Sun Microsystems and IBM/Sun, respectively. Wikipedia+1
Originally they aimed to improve graphics for PC gaming, but over time the company pivoted toward high‑performance computing and artificial intelligence.

“People are going to use more and more AI. Acceleration is going to be the path forward for computing. These fundamental trends, I completely believe in them.” — Jensen Huang

TSMC — Taiwan’s foundry and the silent giant

TSMC (Taiwan Semiconductor Manufacturing Company) was founded in 1987 by Morris Chang, a veteran of the U.S. semiconductor industry. He earned advanced degrees at MIT and Stanford, held senior roles at Texas Instruments, and by 1983 felt his prospects there had stalled. Wikipedia+1

Chang accepted the invitation from Taiwan’s government to lead the Industrial Technology Research Institute (ITRI) and redirect his career to Taiwan. There, he launched TSMC as the world’s very first “pure‑play” foundry—a company that would not design its own chips but focus entirely on manufacturing others’ designs. tsmc.com+1

He described the model in his own words:

“It works, the learning curve, the experience curve, it works only when you have a common location … Learning is local.” Business Insider
And likewise:
“Without strategy, execution is aimless. Without execution, strategy is useless.” Goodreads+1

From the start, Chang insisted that TSMC would “not compete with our customers” — a line that became the company’s guiding mantra. LinkedIn

Under his leadership, TSMC took on the manufacturing burden for many of the world’s leading chip‑design firms. The company made advanced nodes (5 nm, 3 nm) viable, enabling customers such as Nvidia and Apple to push performing devices and AI infrastructure forward. Wikipedia

Chang reflected years later:

“The highest degree of success in 1985, according to me, was to build a great company. … I happened to achieve the highest degree of success that I had in mind.” The Wall Street Journal

In short, while TSMC may operate behind the scenes, its founder and business model changed the global semiconductor industry—and quietly became indispensable. The company may not design the chips, but many of the chips you rely on were made under Chang’s foundry design.

The Symbiosis: Design Meets Manufacturing

In the mid‑2000s, Steve Jobs was quietly working through one of the biggest pivots in consumer electronics: moving from the iPod to the smartphone. He knew he needed a tiny, powerful chip that could deliver performance, battery life and sleek design. Big legacy chip‑makers balked. The story is famously told: Intel turned down Apple’s request to build a chip small enough for a new “phone‑plus‑iPod” device.ationship is the hidden engine of the AI wave.

What Jobs did was lean into partners who could deliver high‑volume, cutting‑edge manufacturing, which helped make the first‑gen iPhone a reality. As one piece puts it: “Many of the chips ‘Designed by Apple in California’ are now manufactured exclusively in Taiwan.”
Fast‑forward to today. NVIDIA does the heavy lifting on chip design for artificial‑intelligence (AI) workloads — massive, power‑hungry, performance‑driven pieces of silicon. Meanwhile, TSMC manufactures those chips, using processes only a handful of companies worldwide can match. When NVIDIA designs something bold, TSMC brings it off the wafer line.

Without NVIDIA’s bold architectures, there would be no new nodes. Without TSMC’s manufacturing scale and advanced node leadership, those architectures would remain drawings. Together they power the AI‑hardware wave, which began as Jobs’ quest for the perfect phone chip has evolved into a global design–manufacture partnership worth trillions. Now, NVIDIA and TSMC have a synergy that goes beyond their $4 trillion evaluation to ensure they meet today’s and tomorrow’s demand.

Manufacturing Leadership & the China Gap

TSMC leads global manufacturing with nodes such as 3‑nanometre and plans for 2‑nanometre, placing it years ahead of many competitors. Meanwhile, China, despite strong ambitions and large investment, still lags in the highest‑end foundry capability. That gap is not just about spending—it is about ecosystem, equipment, experience and global customer relationships. For a general reader: when you hear “AI chip boom,” most of the action is happening in a few advanced fabs in Taiwan, not everywhere.

Risks & the Silent Layer

Behind the headlines of booming valuations lies a fragile supply‑chain reality. Building new fabs costs tens of billions of dollars, takes years and depends on specialized equipment and rare materials. Geopolitical tensions, export controls and concentration of manufacturing add risk. On the environmental side, pushing for smaller nodes means more energy, water and complex logistics. Meanwhile, smaller players and regions risk being left behind entirely, increasing global technical divides.

Why It Matters to You

When you use an AI assistant, stream a video, or interact with a smart service, the hardware powering it likely passed through this design‑and‑manufacture duo. Recognizing who makes the chips and where they are made helps you understand future risks in tech, geopolitics and business. This isn’t just about fancy algorithms. It is about the machines that run them—and the locations that build those machines.

Key Takeaways

• NVIDIA was founded in 1993 and evolved from graphics to AI‑accelerator design.
• TSMC was founded in 1987 and pioneered the foundry model, manufacturing chips for others at scale.
• Their partnership—design by NVIDIA, manufacturing by TSMC—is central to the hardware side of the AI revolution.
• China is working hard to catch up in chip manufacturing but remains behind in the highest‑end foundry nodes.
• The hardware supply chain is a strategic asset and a vulnerability—not just for tech companies, but for global business and geopolitics.

Frequently Asked Questions

Why is NVIDIA so valuable?

• It designs the world’s most advanced AI chips.
• Its GPUs power nearly every major AI model today.
• The company captured a dominant position in a fast-growing, high-demand market.

What exactly does TSMC do?

• TSMC manufactures chips for other companies (like NVIDIA and Apple).
• It uses cutting-edge processes at a massive scale.
• It’s the world leader in semiconductor fabrication, especially at the smallest “nodes” (like 3nm).

Why can’t China just build the same manufacturing capacity?

• Catching up in chipmaking isn’t just about money.
• China lacks access to the most advanced tools (like EUV machines).
• It also lags in talent, ecosystem experience, and trusted global customers.

What are the risks to this hardware supply chain?

• Manufacturing is concentrated in a few regions (mostly Taiwan).
• Geopolitical tensions, especially around Taiwan and China.
• New fabs cost tens of billions and take years to build.
• Supply chain fragility and environmental challenges.

Why should a general reader care?

Knowing where they come from—and how they’re made—offers insight into the future of tech, politics, and global risk.

Every AI tool, smart device, and digital service you use depends on these chips.