This newsletter series will focus on Decision & Position. I often wonder whether my decisions and life positioning are correct, and I frequently observe how people around me make decisions and choose their life positions. Therefore, this series of articles will explore others' decisions and positioning. Through listening to and analyzing others' decisions and positioning, I hope to help optimize my own decision-making and find my own position in this vast world.
I hope that through this newsletter series, I can get to know your decisions and positioning.
Today's content I want to share is a decision I really like and find extremely brilliant—the story of NVIDIA's decision-making during its early startup phase when facing bankruptcy.
This year, NVIDIA's market value has jumped to the top five in the United States, officially joining the trillion-dollar club (1 followed by 12 zeros), and Jensen Huang's personal net worth has reached one trillion New Taiwan dollars. NVIDIA's products span various fields, including AI, genetic engineering, bioengineering, climate, mining machines, gaming, 3D computing, data center services, and more.
But did you know they once faced imminent closure?
NVIDIA once nearly went bankrupt because they couldn't manufacture products that the market needed. At that time, they made a key decision that allowed NVIDIA to successfully survive, and four years after this decision, they successfully went public.
What kind of key decision allowed NVIDIA to successfully survive?
1993: NVIDIA's Founding and Early Products
Back in 1993 when NVIDIA was just established, there weren't many competitors in this market, and they successfully raised $2 million from investors (equivalent to 60 million Taiwan dollars). After two years of preparation, NVIDIA launched their first product, the graphics chip NV1, in 1995. The NV1 had a complete 2D/3D core and also built-in audio processing core. At that time, the market thought NV1 was an interesting product but not worth buying. However, developing this product had already consumed considerable funds for NVIDIA.
NVIDIA invested in developing their second product, NV2, but halfway through development, they faced Microsoft's pincer attack. Microsoft developed a completely new 3D standard (DirectX) by acquiring a startup company, and this standard was completely incompatible with NVIDIA's developing products.
Microsoft's 3D standard (DirectX) was a significant blow to NVIDIA because game developers and the ecosystem were quite resistant to products that didn't support Microsoft's standard. However, NVIDIA discovered that they had burned through most of the company's funds developing previous products, and the company only had 9 months of funding left.
Jensen Huang faced two very difficult choices at this time.
The First Difficult Choice: Abandon Past Efforts and Start Over? Or Go Head-to-Head with Microsoft?
At this time, NVIDIA faced a choice: should they abandon their past product development and efforts, and invest in Microsoft's new standard? If they chose to abandon their past efforts and invest in Microsoft's new standard, dozens of companies were already using Microsoft's standard for development. NVIDIA would transform from a market leader to the last place in the market. They would need to start developing their own technology from scratch, catch up with other competitors, and then defeat these competitors. This would be quite difficult, especially since the company had already burned through its funds.
But if they chose to continue developing their own past technology, they would go head-to-head with Microsoft's ecosystem. When Microsoft launched Windows 95, it was a revolutionary product—the first successful consumer-oriented computer operating system. Microsoft had already captured most of the PC market and developers at this time. If they chose to develop their own technology, NVIDIA might go head-to-head with Microsoft's ecosystem and all developers, potentially losing this ecosystem war and ultimately burning through their funds and going bankrupt.
If it were you, what would you choose?
Jensen Huang said the night of making this choice was an unforgettable night. The company discussed internally for a long time because both paths could ultimately lead their company to failure and bankruptcy. Choosing to join Microsoft's standard and abandon past development results was like denying everything they had done in the past, telling everyone that everything they had done was a mistake. But insisting on their past decisions seemed like a dead end...
Jensen Huang said the company's internal consensus was that choosing a wrong direction cannot lead to the right destination. So Jensen Huang decided to abandon past development efforts, trust each other, and develop chips according to Microsoft's standards.
The Second Difficult Choice: With Only 9 Months of Funding Left, Should They Continue Development? Raise Funds from Investors? Or Dissolve?
When Jensen Huang decided to join Microsoft's standard and redevelop, the company only had 9 months of funding left to survive. At that time, the market average for developing a chip was two years. Developing a chip required two years because they needed to manufacture the chip first before they could write software that could run on the chip. After manufacturing the chip and writing the software, they could then test the chip. If bugs were found in the chip, they needed to remanufacture the chip. And once they remanufactured the chip, they needed to wait for chip manufacturing again and rewrite the software programs. There would be many rounds of testing in between, and this development process took about two years. Therefore, from chip production, software writing, to official mass production required two years.
With only 9 months of funding left, what should they do?
If it were you, what would you choose?
Jensen Huang made a very bold decision: compress chip development time from "2 years" to "6 months." How could they compress chip development time? Jensen Huang discovered that there was a startup in the market manufacturing a type of simulator that could allow software to simulate running on a chip without actually manufacturing the chip. However, this machine was quite slow, and purchasing this machine would use up 3 months of the company's funds.
So Jensen Huang decided to buy this machine, then use 6 months to develop the chip, and told employees: "We don't have time and money to test the chip. We need to ensure that the chip we manufacture is perfect before sending it to mass production. Because we only have one chance. The chip is manufactured, software is loaded, without actual testing, it needs to be sent to mass production. The industry average is two years to complete a product, and we need to complete this product in six months, and it needs to be perfect because we only have one chance."
Finally, they sent this six-month chip (RIVA 128) to mass production. After mass production, NVIDIA's test results showed that this was a chip that could operate normally. Although it wasn't a very perfect product, according to Jensen Huang, this chip could only support 8 of the 32 graphics rendering technologies of Microsoft's standard. NVIDIA worked hard to convince developers that these 8 technologies could complete almost all the animation effects you need, and RIVA 128 was cheap enough and had good enough performance. Even though it had flaws, its advantages still couldn't be ignored.
Developers bought it.
Jensen Huang brought NVIDIA back from the brink through this imperfect chip (and NVIDIA's way of convincing customers).
Looking Back at the Two Choices
First Choice: Courageously Admit Mistakes
Jensen Huang's first difficult choice was to admit his mistakes, abandon all past development efforts, and transform from a market leader to the most backward manufacturer in the market.
In this story, we can see that Jensen Huang had several factors to consider before making his choice: partners (having excellent partners beside him), industry (he believed this industry would have a bright future), ecosystem (a product needs a broad ecosystem to succeed). After seeing these judgment factors clearly, he courageously made his decision.
Second Choice: Break Industry Conventions and Choose a More Impossible Path Forward
According to the common operating model of current startup companies, when facing company funds running out, there are several common patterns:
- Seeking Next Round Investors: Focus on external investors to raise the next round of funding. Although company product development can proceed simultaneously, the main company focus is on how to raise the next round of funding to allow the company to continue burning money for development.
- Seeking Acquisition Opportunities: If no external investors are willing to invest in the next round of funding, perhaps they can explain to other companies that although the company previously developed products incompatible with Microsoft, the company has strong development capabilities. This could convince other companies, such as game developers or tech companies, to acquire them and continue development under a large company.
- Lamenting Bad Timing and Gracefully Closing the Company: Since the company has successfully developed products, although the results are not satisfactory, they still produced some results. Facing Microsoft's attack, any company would fail and close down. They really just lacked good timing, so they have to dissolve at this time. Excellent engineers can use this time to interview for their next company job.
- The Craziest Option: Compress the industry average 2-year development time to 6 months, using the last development opportunity to decide the company's fate.
This was really the craziest option. The goal was set: "Need to develop a product that the market will buy." With only nine months of funding left, they had to compress the development process to within nine months. So they found a rare simulator in the market, then developed and tested simultaneously, just to control the process within possible time limits.
This decision opened my eyes because most people (including me) would be limited by conventions. Since conventional development process requires two years, we can only develop according to the conventional two-year process. After all, this is the established convention.
But Jensen Huang returned to first principles, challenging and breaking conventions. Is there a way to complete it in a shorter time? If possible, then try it.
Some netizens said Jensen Huang's decision was like buying a lottery ticket and winning by chance.
If we only attribute successful results to the courage of going all-in, this decision seems to be just luck plus effort.
Besides luck and the courage to go all-in, what other factors made this decision successful?
I believe the answer is: Positioning
3D graphics chips were a nascent industry. At that time, Microsoft had just released the desktop operating system Windows 95 (retail version released in August 1995), and personal computers were just beginning to flourish. Sony's first-generation PlayStation was launched in Japan in 1994 and successively launched in Europe and America in 1995. From this era's background, we can see how new the 3D graphics industry was, and in a quite new industry, the acceptance of immature products is quite high. This is also why NVIDIA's products with many flaws still sold well—because these industries were just beginning to develop at that time.
If a startup company now wanted to enter personal computer operating systems or smartphone operating systems, everyone would know this would be extremely difficult. The operating system industry is already quite mature, and consumers, developers, and manufacturers will examine any product with very strict standards.
NVIDIA initially positioned itself in a very early-stage developing industry. Combined with good decision-making, they were able to turn the company around successfully on the brink of bankruptcy. Without correct positioning and only bold decision-making, I believe the chances of the company's turnaround would decrease significantly. This experience deeply influenced NVIDIA's future decisions and positioning.
As for how Jensen Huang talks about NVIDIA's positioning, I will introduce it in the next newsletter.
If you also like this newsletter content, welcome to subscribe or share it with friends and invite them to subscribe together.