Jensen Huang
Zero-billion-dollar markets
Entering before the market exists
Public talks and interviews over the years
He has his own name for how the company chooses. How to stay alive inside a market that does not exist yet.
What actually happened?
He describes the choice in one phrase: we go after zero-billion-dollar markets. Not billion-dollar markets. Markets whose present size is zero, where nobody is competing because nobody is yet buying. CUDA is the case. From 2006 the graphics chip was rebuilt for scientific computing and handed free to researchers everywhere, and the spending looked terrible on the statements for close to a decade, until deep learning arrived and everyone discovered the AI world was standing on his foundations. His explanation: if the market already exists, you are playing a game somebody else wrote the rules for.
A zero market buys the seat where the rules get written
Enter a mature market and the product, the channels and the standards belong to somebody else; you are scoring on their board. Enter at zero and the toolchain, the developer habits and the interfaces are yours to write. What you buy is not a head start. It is the referee's chair.
Early in a category the valuable move is not selling units. It is making your interface the default, so later arrivals build on your ground however good they are.
Getting through it takes two lines of blood supply
A zero market pays nothing for ten years, so what keeps the company alive? Two tracks. Gaming cards as the mature business feeding cash, CUDA as the zero track burning it. Without the first the second is suicide; without the second the first is merely a good business.
Feed one dark line off the cash cow and exempt it from revenue targets for five years. Most companies fail this on the quarterly question, not on the money.
Which zero to pick is read off physics, not off research
A zero market has no users to survey, because the demand is not there to ask about. His basis was harder: the growth curve of demand for computation, the physical advantage of running things in parallel. The only dependable map here is where the technology curve points.
Surveys cannot ask about the unseen. Watch where the underlying cost curve is falling and what crossing a threshold would unlock. That crossing is the market's birthday.
How do I use it today?
Where you are: a direction with no market today, and every projection tells you to drop it.
Ask first: if it does come true, who writes the rules? Do I have a line of supply that lasts until then? Am I betting on an opinion or on a physical curve?
Where it goes wrong: treating everything nobody is doing as a zero-billion gold mine. Most markets that are zero stay zero permanently.
Lines to keep
We go after zero-billion-dollar markets.
There are no competitors there, because there are no customers yet.
If the market already exists, you are playing somebody else's game.
Same situation, other people are asking
It's live and nothing is happening. Do I market it or go talk to people?We hit the numbers and I can't say who we serve.There's no market for this yet. How long can I fund it?Everyone still using it says they like it. Who am I not asking?I'm building something nobody asked for · all 4 questions →If this one named what you are going through,
send it to someone who needs it, or keep it somewhere you will find it again.