The startup case study of nVidia that every engineer needs to know.
I wish to tell the most complete version of the story of nVidia, both technically and financially. The story is not amazing because of how successful it becomes, but because of the trouble it had to endure to get there. The success is uncommon, but the troubles are common.
In June 2023, the company nVidia reaches a capitalization of 1 trillion dollars. This is twice as high as TSMC, and four times as high as INTEL. This means nVidia can buy TSMC and Intel. However, such success took 30 years and harrowing trial and error. The company was nearly killed at least six times. Each time it was saved by the wise decisions and luck. It is typical of any startup company, let alone such a great unicorn. Let’s review them one by one
In this world there are unsung heroes everywhere. They are called entrepreneurs, and they build businesses. These are their stories, and how you can become one yourself.
A smart person is not a business builder or founder. A CEO is not a founder, neither. Founder skill is not taught anywhere. It must be self taught in real life. Once you are a founder, you are always a business builder. Here are stories of some founders.
nVidia is a company that makes a special kind of computer chip. The chip can be used both for computing and graphics display. It has an unique architecture that makes them very fast for crunching numbers. Initially the company makes graphics cards to refresh computer game images. Later, the graphic computing chips are used for bitcoin mining, automotive self driving, and artificial intelligence. In this video, we will tell the story of nVidia and founder Jensne Huang. We will intentionally tell the story twice. The first time you will hear the ordinary story that even ChatGPT can summarize for you. The second time you will hear it from a technology and business perspective as nVidia faced off with competitors and rivals.
First, the story on paper.
The Huang family moved to the United States when he was nine. He and his brother came two years ahead of their parents. He is an academic star – entering college at age 16 and majored in hard science. He earned a bachelor’s degree in Electrical Engineering from Oregon State University in 1984, and a Master’s degree in Electrical Engineering from Stanford in 1992. Even before he graduated, he was already working in industry and honed his talents. He was an engineer at AMD and then worked as a director at LSI. nVidia started on his 30th birthday, in 1993. He has to quit a cushy job to start the company and embrace new learning and uncertainty. The company went IPO in 1999.
What makes this one different is the continued growth after the IPO event. In order for a million to become a trillion is 1000 times harder. Thirty years after its founding, the company has reached a market valuation of 1 trillion dollars. Jensen’s 86 million shares of nVidia stock makes him worth 36 billion dollars.
But he won’t need to touch any of the stocks. Jensen’s annual salary is 25 million dollars. That would it 96000 dollars per work day, and 12000 dollars an hour. Is this salary worth it?
Here are some trivia facts. He worked at a Denny’s restaurant as first job. This actually was supremely important for his transformation from an engineer to a manager and salesman. He learned first hand how to communicate and fulfill orders.
Now let’s hear the nVidia story again, with a little more depth. We will show the stock price of nVidia to connect technology events with financial milestones.
nVidia had to cross six hurdles on the journey from zero to 1 trillion.
Hurdle one: the unlikely start
Jensen Huang was invited to join as a key cofounder by his close colleagues. His cofounders are industry veterans of more seniority than him technically. Curtis Priem and Chris Malachowsky are both terrific engineers and chip designers in their own right. But they wanted Jensen to join the team.
In fact, they meet often in a Denny’s restaurant. This is where Jensen Huang did a lot of market studies. But Jensen Huang was the last one to quit his job and join. Why? By the time nVidia started, the field was not empty, but completely full. Large companies like SUN and Silicon Graphics were the Apples and Google of that era. The new startup would have to compete directly with these flag ship companies.
When Jensen asked people if they should start the company, all with intimate knowledge told him no. When they reach the investors, no one supported them. To make it worse, although Huang was a high level managers now, he has never started companies before. He struggled to write a business plan, but never finished it.
After six month, two investors believed in his team and gave them funding to start. However, the confidence is not in the technology, but in the team. For investors, the investment is a bet on a horse, not the idea. Many people have the same idea. Many big companies are in the field and market.
Even the investors would be surprised today. They bet on the right horse. Were they sure about the bet? I am sure the answer is no.
Hurdle two: Making break even.
The founders have worked as employees in other companies, but they have never had this much responsibilities before. nVidia’s first product was bad. Second product failed. However, these are all within the plan.
The first product failed badly. It performed OK but the price was high. They have to cut personnel from 100 to 30 to save cash. In fact, the product was only made with 30 days left in the company’s cash life line. At this point, the situation seems bleak.
However, there was a lucky break. The company caught the attention of SEGA, a Japanese electronics game maker. SEGA was mired in a market share battle with Nintendo, and both are trying to make games with high resolution and 3D features. SEGA gave nVidia seven million dollars for a contract project, and the money supported the company’s pay roll. Seven million dollars was a lot of money in the nineties. Without SEGA’s money, nVidia would have to beg for VC money continuously and dilute shares.
However, nVidia eventually realized that they were following the wrong technology path on the SEGA project. Huang was in a dilemma. If he told SEGA the truth, they can pull the funding and the company would close door and fire everyone. If he does not tell SEGA, the company will be making a wrong product and it will eventually die a few months later, only in a deeper hole.
For smart people, humility is hard to teach. Jensen Huang is smart enough to be humble and honest with himself. He talked with SEGA’s project leader and, to his surprise, SEGA gave him the full payment and pulled the plug on the project. No longer obligated to finish the project, nVidia got a new life by sheer luck and generosity.
No good deeds go unpunished. Sega could not release the intended product, SONIC Xtreme. Quickly, Nintendo was able to grab a market share. Ironically, Nintendo choose to work with Silicon Graphics to build the chip. The Nintendo 64 becomes an iconic system and gaming platform for decades.
Hurdle three: faced off with first rival
As we mentioned before, there are many similar companies. Each one led by smart people and funded by a lot of money. They all engage in hand to hand combat, trying to survive. Being smart is good, but it is not enough.
As Jensen said himself later,
But that is not enough.
By this time, Jensen Huang and his team has been beaten twice, and the company has not closed its door yet. The story of nVidia has not even began. What made nVidia, a two time loser, to win the next battle?
Without a successful product, the company will never be able to reach break even point. It would forever need new investment money or project money. That is not sustainable.
Huang has been secretly building another arsenal – the talent pool and managers/administrators. They will come handy very soon. As a founder, it is both important to build the business and the company. Jensen never went to MBA school, but somehow he is good at building company and culture. This is likely what the investors saw as unique.
Jensen said in an interview, “As founder CEO, selecting people is 99% of your work”.
At this point, Jensen Huang has at least three major and powerful competitors. The Microsoft. 3dfx. And ATI. All of them formed by well educated men and teams. All small companies wanted to fight Microsoft and introduce their own standard. However, nvidia decided to embrace Microsoft’s standard and work with Microsoft. As such nVidia get to have an allay, since Microsoft does not want 3dFx to have monopoly.
However, 3dfx’s graphic card product is very powerful. The team was founded by veterans from silicon graphics. For nVidia to build a rival product, they must be both fast and good. They only have one shot to make it right, or the curtain will close. Huang has a tough decision to make. To buy time, they have to use 3 month worth of money to support a new simulation company and their untested simulator. He took the risk and pulled the trigger. It was a gamble that paid off. Their first and only design of a complex circuit worked. The new chip started to sell, and the company is not short on money for the time being.
The reward for nVidia is to buy 3dFx and eliminate a rival. And the company is listed as a public company, finally starting its journey on the big board.
For the first time, the general public can buy nVidia stock now, and the investors and founders can now sell their stocks to get rich. 1999 was a good year, just ahead of the impending tech bubble burst.
You can see on the stock price chart of nVidia the rise and fall around the time 3dFX was acquired.
Hurdle four: a new rival arrived on scene
The next challenger is a company that is even older than nVidia. A Canadian company called ATI is a serious competition. It acquired a silicon valley company called ARTX in 2000, which are formed by 20 veterans in SGI who designed the Nintendo 64 products, which beat nVidia and SEGA earlier.
Let me side track to discuss a bit about ARTX. The company was founded by a number of industrial veterans, among them is Wei Yen. He rose to become a vice president of SGI, but quit his job to start two companies before the burst of tech bubble. In terms of startup success and experience, he is on par with Jensen. In a parallel universe, he could become Jensen Huang.
This time, nVidia and ATI are both equally good. They traded punches. Anyone can win. But the company with the best people and culture won. ATI lost core business, and was acquired by AMD. AMD seriously wanted to compete with nVidia now, as well as Intel.
After beating 3DFX and ATI, nvidia now will face even more worthy opponents. It is in the big league.
You can see the jump of nvidia stock after ATI was acquired. But this euphoria did not last long. In fact the stock price will hover between 4 to 10 dollars for another ten years. The reason is because the company has a hidden enemy it must beat.
Hurdle five: beating myself
After many years of growth, you would think the company will be in no trouble. You would think it is time to collect passive income for life. You would be wrong. It turns out that the total market of graphics card is limited. If the company only becomes the king of graphics card, it will eventually have to slow down. By slowing down, it is just waiting for young companies to take over. In business, you stop, you die. The nVidia team must pivot into bigger markets. For that, it must invade the turf of Intel.
The company’s biggest rival is now itself. It must startup a second time, build a different product and take on more uncertainty. For that, Jensen’s team built a product that no one has done before and no one want – the GPU.
The GPU is a rival to the CPU. However, if the CPU is a big brain, the GPU is a collection of smaller muscular brains. Not only did the product take a long time to develop, it also faces a long struggle to sell. People just never demanded this, no matter how useful it is.
You can see from the stock performance, that the company got punished in the stock market. Its stock was flat for a decade. Fortunately at this point the company always have enough revenue to pay for employees and engineers.
It took nearly 10 years for the company to catch some favorable wind. In 2016, the stock ticker of nVidia finally started to move because of the decision made always 15 years ago. The artificial intelligence community was motivated to use the GPU to solve their problems. The people who mine the bitcoin use their chips. The stock of nVidia rose like a bubble. But is it going to last?
Hurdle six: Reaching trillion by ditching weight
In early 2023, the bubble of bit coin was serious dashed by governments. Naturally, the company has to do some amazing act to reach a valuation of 1 trillion. It was tagged by Google to make Android phone chips. If they can get this market, the sales will be unbelievably high. However, the competition is fierce. If nVidia compete, it may win in a dragged out battle, or may lose eventually. The mobile phone chip development would tie up a lot of internal resources.
Jensen Huang’s team decided to abandon the Android phone market. Instead, it started to focus on a nascent new area of self driving cars. Their chips become a standard brain for the operating system of self driving. In 2023, the long awaited but never expected success of Jensen Huang finally arrived. The company’s stock price reached a historical high, following the public’s interest in Tesla self driving cars and ChatGPT.
You can see on the chart the boom of stock price due to bitcoin mining, and then the sudden collapse. Fortunately, ChatGPT came to the rescue in 2023. For nVidia, the journey is still on going. Business will test Jensen Huang everyday, to navigate the choppy water of tech.