Business9-16

The Invisible Giant Inside Your Phone

There's a Name in Your Pocket You've Never Heard Of

Pick up your phone. Or your game console. Somewhere inside that thing is a chip about the size of your thumbnail that does the math, draws the picture, and connects you to the internet. Who actually made it?

You'd probably guess Apple. Or NVIDIA. Or whichever big-name company you've heard of.

Here's the twist: almost none of those companies actually make chips. They design them — they draw the blueprints. The company that turns those blueprints into real, physical chips, one silicon wafer at a time, is a name you've almost certainly never heard. It sits on the island of Taiwan, barely advertises, almost never makes headlines — and yet its chips end up inside your iPhone, your game console, your dad's car, and even fighter jets.

That company is called TSMC (Taiwan Semiconductor Manufacturing Company). In 2021, the world ran so short on chips that Ford — yes, the car company — had to stop making its best-selling truck, the F-150. A truck needs dozens of chips, and missing just one meant the whole vehicle couldn't be built. The hosts of the podcast Acquired put it in one heavy sentence: "Semiconductors run everything, and they run semiconductors."

Here's the part that makes this story even stranger: the man who built this company the entire world can't live without started at age 56 — and he'd just been pushed out of not one, but two jobs. This isn't a story about some hotshot twenty-something coding in a garage. This is a story about a man everyone considered washed up, pouring every ounce of experience he had left into growing something enormous.

Let's start from the beginning.

Chapter 1: The Boy Who Fled Three Times

In July 1931, a boy was born in Ningbo, China. His name was Chang Chung-mou — you'll know him better by his English name, Morris Chang. His father worked for the county government and later became a bank manager; the family was reasonably comfortable.

But his childhood was anything but stable. Before he even turned 18, he had to flee his home three times. In 1937, he escaped with his mother to Hong Kong. On December 8, 1941 — just three hours after Japan bombed Pearl Harbor — Japanese forces invaded Hong Kong in only three hours, and the family fled back to Shanghai. In 1948, civil war broke out in China, and they fled to Hong Kong once more. By the time he turned 18 and got ready to leave for America, he had packed up his whole life three separate times, bounced between two cities, and watched a stable home get shattered overnight by war — not once, but repeatedly.

At 18, with help from an uncle living in Boston, he got into Harvard. He later described the moment he walked through Harvard's gates:

"My reaction entering Harvard was sheer ecstasy, almost disbelief. What a country!"

But the excitement didn't last long before he coldly sized up his situation. In the early 1950s, there were only a handful of "respectable" careers open to Chinese immigrants in America — laundry owner, restaurant owner, engineer, or professor. Not lawyer. Not accountant. Definitely not politician. That was the career ceiling Chinese immigrants faced back then. And Harvard didn't even train engineers — so he transferred to MIT, switched to mechanical engineering, and finished both his bachelor's and master's degrees in just three years.

Notice something: this wasn't the last time he'd look hard at the cards he'd actually been dealt and pick the one path that could actually work. That exact way of thinking would resurface 35 years later — and change the entire semiconductor industry.

Chapter 2: The Fork in the Road That Cost $1

Morris failed his MIT PhD qualifying exams — twice, which was the school's limit. He later joked about it dryly: "Unfortunately I did not pass the qualifying exam; fortunately they were kind enough to let me take it again — I failed that one too."

With a PhD off the table, he needed a job. Two companies made him offers at the same time: Ford Motor Company offered $479 a month — his dream job at the time. A company called Sylvania offered $480 a month for its semiconductor division. A one-dollar difference. He asked Ford to match it. Ford said no.

Because of that single dollar, the auto industry lost a mechanical engineer forever — and the semiconductor industry gained Morris Chang.

During his three years at Sylvania, he had zero background in electrical engineering and had to teach himself everything. At night, he'd read a 1950 electronics textbook, and whenever he got stuck, he'd head down to the hotel bar late at night and buy drinks for a senior engineer who liked to drink, in exchange for answers. He later said:

"He didn't solve all my problems, but he solved enough so that I could move ahead. He was my main teacher about electrical engineering."

What finally pushed him to leave Sylvania was a speech from a company executive: "We cannot make what we can sell and we cannot sell what we can make." The moment he heard it, he knew this ship was sinking, and he got off.

Quick decode: why is making a chip so brutally hard? Picture "making a chip" like baking the world's most demanding cake — except you have to draw lines thousands of times thinner than a human hair, on a space smaller than your fingernail. A speck of dust or a tiny temperature swing, and the whole batch is ruined. The first time a factory tries a brand-new recipe (a new generation of chip technology), the success rate — called the "yield" — often starts terrifyingly low, sometimes just a handful of usable chips out of every hundred tries. The only way to raise it is to keep baking, keep tracking what went wrong, and keep adjusting. The more batches you run, the faster you learn, and the higher your yield climbs. That's exactly why chip factories dread having too few orders: fewer batches means less learning, which means the yield climbs painfully slowly.

Chapter 3: Thirty Years at TI — From Star Player to "Put Out to Pasture"

In 1958, Morris joined Texas Instruments (TI), then the biggest semiconductor company on Earth. That same year, by coincidence, two other engineers independently invented the integrated circuit — the ancestor of the modern chip.

His first big project was to be a "second source" supplier of chips for one of IBM's mainframe computers — basically, a backup in case IBM's main supplier fell through. IBM's own production line only hit about a 10% yield. TI's line started at 0%. Using his engineering instincts, Morris pushed that yield from 0% to 20% in about four months — double what IBM managed on its own line. That win landed him his first management job and put his name on the radar of TI's top brass.

His second big win was even more impressive. Around 1967, working with a then-small consulting firm called BCG, he flipped the entire industry's pricing playbook upside down. The standard move back then was to price new chips high, to recoup the enormous cost of building the factory. Morris did the opposite — he priced new products low from day one, and then automatically cut prices every single quarter, even when nobody was asking for a discount. He later said:

"A lot of people thought we were being foolish. Why would you reduce the price when you didn't have to? But we did it because we believed in it, and indeed our market share just kept expanding."

Here's the logic: the higher the price, the fewer people buy, which means the factory runs under capacity, which means less practice, which means the yield climbs slower, which means costs never come down — a vicious cycle. Flip it around: lower prices bring in more orders, the factory runs flat-out and "practices" constantly, yields climb fast, costs drop fast, and profits actually go up. That single move turned TI's chip business into the biggest and most profitable in the world.

Everything was going great — by 1972 he'd been promoted to Vice President, running the whole semiconductor division, one step below the top job. Then it all turned. In 1978, the company shifted him over to run its money-losing consumer products division — a fix that never came, not in five years. In 1983, he was demoted to a job with a deliberately harmless-sounding title: "head of quality and people effectiveness." In his own words, he'd been "put out to pasture." At 52, he resigned.

Quick decode: what's actually clever about "learning-curve pricing"? Imagine you just started baking pies. Your first pie takes an hour and often comes out burnt. The more pies you bake, the better you get — after ten tries, maybe you can make a beautiful, delicious pie in fifteen minutes. Here's what Morris figured out: if you price your pies sky-high right out of the gate, hardly anyone buys them, which means you barely get to practice, which means your skills never improve. Better to price them cheap on purpose, so lots of people buy — and while you're filling all those orders, you're getting faster and better, and by the time your costs drop way down, your competitors are eating your dust.

Chapter 4: Another Setback, Then Taiwan at 55

After resigning, Morris moved to New York to become second-in-command at a company called General Instrument. He lasted exactly one year — he decided the place operated more like a company that bought and sold tech assets for profit than one that actually built things, and that clashed with his own values of doing real research and real engineering. He later admitted that after two setbacks — at TI and here — he no longer believed he had a shot at becoming CEO of a major American company.

In 1985, a Taiwanese government official named K.T. Li invited him to lead Taiwan's Industrial Technology Research Institute (ITRI), nicknamed "the Bell Labs of Taiwan." Every former colleague around him told him not to take it. He didn't expect much from it either — he treated it more like an early retirement.

Not long after Morris arrived at ITRI, K.T. Li gave him a direct order: don't hand this to anyone else — you, Morris Chang, are going to found a brand-new semiconductor company in Taiwan, and you're going to make it a world leader. He was first told he had a week to write the business plan. A day later, that got moved up to Friday — just three days. Morris later described it this way: "It was like in the movie The Godfather. It was an offer I couldn't refuse."

In those three days, he took a hard, honest inventory of everything Taiwan actually had going for it — and arrived at what became the single most important paragraph in the entire company's history:

"We had no strength in research and development... no strength in circuit design, IC product design... little strength in sales and marketing, and we had almost no strength in intellectual property. The only possible strength in Taiwan that we had, and even that was just a potential one, not an obvious one, was semiconductor manufacturing, wafer manufacturing. And so what kind of company would you create to fit that strength and avoid all the other weaknesses? The answer was a pure-play foundry."

TSMC was officially founded in 1987. The company raised $220 million total: half from the Taiwanese government, 28% from the Dutch company Philips, and the rest scraped together when the Premier of Taiwan personally leaned on local business leaders to invest. Here's the strange part: Morris himself walked away with zero shares. The company's founding valuation was set at exactly $0.

The entire industry laughed at the idea. A rival chip executive of the era, Jerry Sanders, coined a line that stuck: "Real men have fabs." Translation: if you don't build your own factory and only make chips for other companies, you're not a serious chip company.

Quick decode: what is a "pure-play foundry"? Picture a giant shared commercial kitchen. Back then, every chip company had to do everything itself, from designing the recipe to buying its own oven and mastering its own baking technique. TSMC's idea was different: we won't design the cake, and we won't put our name on the box — we'll just become the best shared kitchen on Earth. You bring the recipe, we'll bake it faster, better, and cheaper than anyone else. That meant small companies who were great at designing cakes but couldn't afford to build a kitchen could suddenly bake their cakes too — and sell them to the whole world.

Chapter 6: Surviving on Leftovers

TSMC's early years were rough. Big chip companies only handed TSMC their orders when their own factories were too busy, or when they wanted to dump a product line they no longer cared about — and the moment their own capacity freed up, they yanked the business right back. Morris himself described it as anything but a stable market — TSMC was living on other companies' leftovers.

But Morris saw something nobody else had spotted. During his years at TI and General Instrument, he'd watched countless talented chip designers who wanted to strike out and start their own companies — but almost none of them could raise enough money to build their own factory. He said:

"When I was at TI and General Instruments, I saw a lot of integrated circuit designers wanting to leave and set up their own business, but the one thing or the biggest thing that stopped them from leaving those companies was they couldn't raise enough money to form their own company... What very few people saw, and I can't tell you that I saw was the rise of the fabless industry, I only hoped for it. But I had better reasons for hoping for it than the people at Intel, TI, and Motorola because I was now standing outside."

Sure enough, from around 1989 to 1993, Qualcomm, Broadcom, Marvell, and eventually the now-famous NVIDIA (founded 1993, raised a total of just $20 million, and never built a factory of its own) were all born inside TSMC's shared kitchen.

Quick decode: the chip industry as one giant class project Picture the entire chip industry as a class assignment split into jobs: some students draw the blueprints (companies that license "chip architecture"), some pick the right tools (companies that build the machines), some turn the blueprint into an actual product and sell it (Apple, NVIDIA, Qualcomm — "fabless" companies that design but never build factories), and TSMC handles the hardest, most expensive job of all: actually baking the chip. Back when everyone tried to do everything alone, it was slow and expensive. Once the work got divided up, everyone specialized in what they were best at, and the whole project got faster and better. That's exactly why a swarm of broke, factory-less young companies could launch successful businesses armed with nothing but a blueprint — because TSMC had already agreed to carry the hardest part for them.

Chapter 7: A Dollar Short — How Intel Lost the Whole iPhone

Before we get to how TSMC landed Apple, there's another story worth telling first — because it's the reason TSMC eventually got a shot at the biggest deal of its life.

In the mid-2000s, Apple founder Steve Jobs was preparing to launch the very first iPhone, and he approached the CEO of Intel — then the world's biggest chipmaker, already supplying processors for Apple's computers — hoping Intel would make the iPhone's chips too. But Intel's CEO decided Apple's offer was priced too low, and turned it down.

Apple went elsewhere for its chips, and eventually bought its own chip-design company to design its own silicon. But no matter who designs a chip, someone still has to build it — and years later, the company building those chips turned out to be TSMC. The hosts of Acquired did the math and called it "one of the biggest strategic mistakes of all time" — Intel didn't lose the deal because of bad technology, and not because they couldn't see the future of phones coming. They lost it purely because they thought the money on the table wasn't enough.

Sound familiar? It's the same story as that $1 gap from the beginning — Ford lost Morris Chang over a single dollar; Intel lost the entire iPhone chip business because a number looked too small. The lesson the hosts pulled from it: never walk away from a long-term opportunity you clearly understand, just because you couldn't agree on price.

In 2005, at 74, Morris retired for the first time, handing the company to his longtime deputy Rick Tsai. But four years later, in 2009, at 78 years old, he came back as CEO — because he'd concluded that the personal-computer era was ending, and the wave of mobile phones and cloud computing was just getting started, and he happened to know nearly everyone riding that wave.

The deal that truly launched TSMC into the stratosphere was landing Apple's business. Around 2010, TSMC poured in $9 billion, mobilized 6,000 employees, and built a factory dedicated entirely to Apple in just 11 months. The Apple executive who led the negotiation put it bluntly: "If we were to bet heavily on TSMC, there would be no backup plan." — a bet-the-company move if there ever was one. At first, Apple also used Samsung as a backup chipmaker to spread the risk, and it wasn't until the iPhone 6 launched in 2014 that Apple gave TSMC the business exclusively.

Here's the interesting part: even after landing such a massive deal, TSMC never advertised "we built this chip" anywhere. Every time Apple unveils a new chip at a keynote, all the spotlight goes to Apple's own brand. TSMC is perfectly happy staying invisible, handing every bit of the spotlight to its customers, and quietly doing excellent work behind the scenes. That choice to stay in the shadows is exactly why every chip company on Earth — designers who need someone to build their chips — trusts TSMC with their business.

Quick decode: why do chip factories keep getting more expensive? There's a famous rule called "Moore's Law": the number of components you can pack onto a chip roughly doubles every so often. There's a lesser-known but far more expensive rule too: the cost of building a factory that can keep up with the most advanced technology roughly doubles every four years — today, a single factory costs $15 to $20 billion. Put those two rules together, and it means that chasing the cutting edge gets astronomically more expensive over time, and fewer and fewer companies can afford to keep up. Morris compared the whole industry to "a treadmill that speeds up all the time. If you can't keep up, you fall off."

Chapter 8: A Global Chip Shortage, and a Trillion-Dollar Invisible Giant

The year this episode was recorded, 2021, sat right in the middle of the worst global "chip shortage" in memory — Ford had to stop building its best-selling truck because it couldn't get enough chips. By then, TSMC was already the ninth most valuable company on Earth, controlling more than 90% of the world's most advanced chip-manufacturing capacity. In 2020, TSMC took 85% of its profits and immediately poured them straight back into building more factories — a flywheel spinning faster and faster: earn money → build a more advanced factory → stay ahead on technology → customers build better products → customers grow bigger and new customers keep appearing → earn even more money.

Years later, looking back, TSMC's market value climbed from $550 billion (when the podcast first covered it) to well over one trillion dollars — making it one of only two trillion-dollar companies in the entire world not located on the west coast of the United States. (The other is Saudi Arabia's oil company, Saudi Aramco.)

Quick decode: why does "doing everyone else's hardest job" turn you into a giant? Over a hundred years ago, every brewery in Germany had to generate its own electricity just to brew beer. Eventually, breweries realized it made more sense to let a power company handle electricity, so they could focus purely on brewing great beer. The hosts made a similar point on the show: whenever you spot a huge crowd of companies all grinding through the same unglamorous, unavoidable, expensive chore — like every single one of them having to build its own multi-billion-dollar factory — that's usually a massive opportunity. Whoever does that chore best, cheapest, and most reliably becomes the wire that everyone else plugs into. TSMC is to chipmaking exactly what the power grid is to breweries.

Quick decode: why can't anyone with money just catch up? You might wonder: if making chips is this profitable, why doesn't some other country or company just spend a fortune and build an identical factory? A journalist on the show put it plainly: "They can spend the money and are, but they wouldn't know what to do with it. It's not because they're dumb." Think of it like this: even if you bought the exact same stove and knives a world-champion chef uses, you still couldn't cook their signature dish — because the decades of instinct, judgment, and muscle memory baked into their hands isn't something money can buy. TSMC's decades of accumulated technical know-how, its relationships with equipment suppliers, and the hard-won experience of thousands of engineers simply can't be bought — or copied overnight. That's exactly what makes this company's moat so unusually strong.

Chapter 9: One Island, and the Whole World Watching

There's one part of this story that deserves to be told to you straight, no sugarcoating.

Nearly all of TSMC's most advanced factories sit on the island of Taiwan. And that creates a genuine problem: there is a long-running, unresolved political disagreement across the Taiwan Strait — between mainland China and Taiwan — and both sides hold their own positions on it. Meanwhile, almost all of the world's most advanced chip-manufacturing capability sits concentrated in this one place, with no backup anywhere else.

The hosts of the podcast consider this TSMC's single biggest risk: if a war ever broke out across the strait, the entire world's chip supply would take a massive hit — and you've already seen a small preview of what that looks like in this very story: a simple chip shortage was enough to halt production of a best-selling truck. Now imagine what would happen if the world's most advanced chips were cut off entirely — nearly everything that runs on chips in our daily lives, phones, cars, hospital equipment, airplanes, would be affected. Even Boeing, the company that builds airplanes, can't build a plane without semiconductors. Because of exactly this risk, the United States has spent recent years pushing for more "technological self-reliance," including subsidizing a new TSMC factory in Arizona — but the hosts point out those factories don't use the most cutting-edge technology, and won't be able to fully replace Taiwan's capacity anytime soon.

The hosts also raised a blunt thought experiment: if a conflict actually happened, TSMC's engineers could evacuate, and maybe the expensive manufacturing equipment could even be disassembled and flown out — but the thing that actually makes this company valuable isn't the machines. It's the decades of technical know-how carried around inside the heads of thousands of engineers. Can that kind of know-how be "airlifted" somewhere else and take root again, the way people and equipment can? The hosts brought up a counterexample: Japan's Toyota once built a joint-venture factory with America's General Motors (the very same factory Tesla later took over to build cars) and tried to transplant its entire production system there exactly as-is — it didn't fully work, even though that attempt happened in peacetime, with zero threat of war. In the end, the hosts didn't offer a tidy answer — they simply admitted, "I don't know." Which is a reminder that this company's toughest, deepest moat is also the part that's hardest to simply "move" or "protect."

This is a complicated, real, still-unfolding situation with no easy answer — and it's worth paying attention to as you grow up.

Some Fun (and Slightly Wild) Facts

TSMC's founding valuation was exactly $0. Morris Chang, the founder, walked away with zero shares. He later built his own fortune mostly by using his salary to buy the company's stock over the years — reportedly ending up worth around $3 billion.
One of the key machines used to make advanced chips fires a laser at molten droplets of tin 50,000 times a second to create a special kind of light. Each machine costs about $200 million, needs four Boeing 747s to ship, and only about 50 of them get built in the entire world each year — TSMC has them all pre-ordered years in advance.
TSMC and ARM — the company behind the chip architecture used in almost every phone today — were founded in the exact same year, 1987. Back then, people in the industry mocked ARM's low-power chip idea: "PCs are always plugged in, so what do we need a low-power chip for? This thing's pathetic." Twenty years later, nearly every phone chip on Earth uses ARM's designs — and almost all of them are built by TSMC.
Texas Instruments, the company where Morris Chang spent 25 years, actually started out in the 1930s building seismic equipment for oil exploration. Decades later, one of the podcast hosts summed it up with a line that stuck: "Semiconductors are the new oil, David."
In the 1980s, the label "Made in Taiwan" mostly showed up on cheap goods — Barbie dolls, toys, clothing — a symbol of low-end manufacturing. Back then, the entire chip industry scoffed: "This isn't manufacturing Barbie dolls here. This is real technology." Three decades later, it's exactly those once-dismissive top chip companies lining up to hand their most advanced chips over to TSMC in Taiwan.
Before Morris Chang ever arrived in Taiwan, the government had already licensed some semiconductor technology from America's RCA and used it to help launch another company, UMC — which later spun off a division that became today's MediaTek, now worth roughly $50 billion. It's one of the very few things the government got right before Morris showed up.
Late in life, when asked in an interview how he'd managed to quit smoking, Morris answered: "I never stopped. I still smoke."

Three Things This Story Teaches You

First: age was never the finish line — experience is the fuel. Morris Chang founded TSMC at 56, after being pushed out of two jobs and even going through a rough patch in his marriage. Nearly everyone around him told him to stop trying. But it was exactly those thirty hard years grinding through the semiconductor industry that let him instantly recognize which cards Taiwan actually held — and spot an opportunity nobody else in the world had noticed yet.

Second: knowing your real strength beats desperately copying everyone else. Taiwan had almost no edge in design, sales, or patents. Instead of trying to out-compete giants like Intel at everything, Morris took the one card Taiwan actually held — wafer manufacturing — and pushed it to become the best in the entire world. Finding the one thing you're genuinely good at, and pushing it to the limit, usually beats trying to be good at everything at once.

Third: a genuinely good strategy often takes years before anyone admits it was right. TSMC's first few years survived on nothing but other companies' leftovers, and the entire industry insisted the "pure-play foundry" idea had no real market. Morris and his team stuck with it for years — until companies like Qualcomm and NVIDIA proved, by actually existing, that TSMC's bet had been correct all along. The ideas that matter most often look like a joke at first.