← Back to videos
今晚别刷 Netflix 了 去看这场黄仁勋16年前在斯坦福讲创业的课,比上两年MBA

Hux
@huxlab
November 16, 2026· 59:33
今晚别刷 Netflix 了 去看这场黄仁勋16年前在斯坦福讲创业的课,比上两年MBA还值。 他刚开始就说,别用“愿景”这种大词。 每个人都有自己的视角,来自自己的经历。 1993年他们觉得PC该能跑3D游戏,当时这个市场是零。 早期最狠的决定是:直接忽略客户。 客户都说芯片太贵,不买。 他反而把公司最后几百万砸进去做更大的芯片。 因为他认为3D图形需求是无限的,摩尔定律会让性能翻倍、成本减半。 结果卖爆了。 后来几百个公司做同样的事,最后只剩NVIDIA。 他说不是因为技术更强,是因为视角不同。 更关键的是,他主动把固定功能的GPU改成可编程的(shader),差点把公司搞死。 但如果不改,行业就到头了。 他反复强调一件事: 创业不是为了赚钱。 钱不是唯一不该创业的理由。 真正能撑过那些几乎把公司搞死的时刻的,是你真的相信这件事,并且愿意反复重做。 很多人学商业学的是框架和案例。 真正拉开差距的,是那种在市场还是零的时候就敢押注、在客户都说不的时候还敢干、在成功之后还敢自己颠覆自己的判断力。 技术会过时,框架也会过时。 这种对“该不该做、什么时候该重做”的直觉,才是最难复制的。 https://t.co/mvVIbDdWqw
Transcript
- 0:02
- Thank you, Tina
- 0:03
- I gave my last
- 0:05
- last pennies to
- 0:06
- Jim Plummer to build that building
- 0:08
- I am now officially broke
- 0:12
- So, instead of giving you a
- 0:15
- company presentation today
- 0:16
- what I thought I would do is
- 0:18
- just have a conversation with you
- 0:20
- you know, at any time if you have a question
- 0:23
- if you would like to change the direction of our
- 0:25
- of our conversation, just raise your hand and we'll
- 0:27
- we'll, we'll talk about whatever comes up
- 0:31
- a lot of people talk about and write about
- 0:35
- building companies
- 0:36
- and I can tell you firsthand
- 0:38
- that building a company is extraordinarily gratifying
- 0:44
- It is also incredibly hard
- 0:48
- And so the things that
- 0:49
- that
- 0:52
- that you want to talk about
- 0:54
- with respect to the company building process
- 0:57
- is rather, expansive
- 1:00
- You could talk about company building processes
- 1:03
- from a lot of different perspectives
- 1:05
- And so I'm going to try to touch
- 1:06
- on a few of them that I think are
- 1:09
- are particularly, important in my experience
- 1:11
- So sixteen years ago
- 1:13
- NVIDIA had three people
- 1:15
- three engineers
- 1:18
- And... Go on
- 1:24
- Did I, did I do something?
- 1:25
- Was it me?
- 1:27
- we had a perspective that
- 1:30
- that if we... I was just talk
- 1:33
- So he, they go
- 1:54
- You think it's me?
- 1:59
- I thought I just did that
- 2:02
- See, it's
- 2:04
- you can't be, you can't be, you can't control user stupidity
- 2:09
- you know?
- 2:11
- You're good, you're good
- 2:13
- It was probably me
- 2:13
- I was sitting on it
- 2:17
- All right, so, so
- 2:19
- sixteen years ago, we started NVIDIA, and, the, the, the insight that we had
- 2:25
- some people call it vision
- 2:28
- vision's an awfully big word to me
- 2:31
- Vision's an awfully big word to me because I
- 2:33
- I believe, first of all, vision matters
- 2:36
- Let me tell you that vision matters
- 2:37
- and I'll help you understand that in a
- 2:38
- second
- 2:40
- but I like to use the word perspective
- 2:42
- because it makes it possible
- 2:44
- for anyone to have one
- 2:47
- When you say vision
- 2:48
- it feels like
- 2:49
- only a few
- 2:50
- selected visionaries of the world can have one
- 2:53
- But everyone has a perspective
- 2:55
- and that's in fact all vision means
- 2:57
- That you see the world
- 2:59
- in a way that is either
- 3:01
- different, or, otherwise, okay, than, than somebody else
- 3:08
- and you see opportunities that I think are
- 3:10
- that you believe are particularly important to go in and address
- 3:14
- that you can address
- 3:16
- in a particular way
- 3:17
- And so perspective, our perspective at the time
- 3:20
- this is 1993, you guys won't, won't remember this, but, the, the PC
- 3:27
- was Windows 3.1, CD-ROM, was
- 3:32
- about to be introduced
- 3:35
- there were no PCs with networks
- 3:40
- Wireless technology, the no, if you said some
- 3:43
- if somebody said radio
- 3:45
- I think you would
- 3:46
- the word that would come to mind is
- 3:47
- FM radio
- 3:50
- and so wireless technology didn't exist
- 3:52
- the fastest microprocessor
- 3:55
- in the world was a sixty-six megahertz forty-six DX2
- 4:01
- and I don't think any of you would even use it
- 4:04
- in your tennis shoes today
- 4:07
- and we would run our computers with that
- 4:09
- and, the PC was, was, becoming used
- 4:13
- for, desktop or for office automation
- 4:19
- our perspective was that this particular
- 4:21
- device was going to be
- 4:22
- unique in the sense that
- 4:24
- it has the ability to run programs
- 4:26
- And what if, we
- 4:28
- gave it the benefit of running 3D graphics programs?
- 4:31
- So that, you could, explore new worlds, play games, you know, play games
- 4:39
- And so
- 4:40
- we started a, a company, and, and the business plan
- 4:43
- basically, read something like this
- 4:45
- We're gonna take
- 4:47
- technology that was available only in the most
- 4:50
- expensive workstations
- 4:52
- We're gonna try to make it
- 4:53
- reinvent the technology and make it inexpensive
- 4:57
- And, the killer app was video games
- 5:01
- and so I, I took this idea to Sand Hill Road
- 5:05
- and, and they told me there was no video game market
- 5:09
- people don't start companies to play games
- 5:11
- And, my parents, my
- 5:14
- I remember calling my mom and telling her
- 5:15
- that I'm gonna start this company
- 5:17
- and she says
- 5:17
- you know, what do you
- 5:18
- what do you guys do?
- 5:19
- And I said, we build these things called 3D graphics chips
- 5:22
- and, and, people would use them to play games
- 5:26
- And, and then she, she said
- 5:27
- why don't you go get a job?
- 5:30
- And, and so, and, and so
- 5:33
- now of course, of course, games was, was, we believe
- 5:37
- going to be a very large part of the marketplace
- 5:39
- Now, we had that perspective
- 5:41
- for very obvious reasons
- 5:42
- We grew up in the video game generation
- 5:45
- I was the video game generation
- 5:46
- I was the beginning of the video game
- 5:48
- generation
- 5:49
- And so the, the entertainment value of video games
- 5:52
- computer games, was very obvious to me
- 5:55
- and I could imagine how it could be
- 5:57
- a very large market
- 5:58
- and a very large industry
- 6:01
- For a lot of the people that were older
- 6:03
- that sensibility didn't exist
- 6:06
- And so notice
- 6:07
- I've just described to you
- 6:08
- a perspective about the world that we had that is apparently
- 6:13
- obviously now true
- 6:14
- because video games is
- 6:15
- the world's largest digital media industry today
- 6:20
- it is apparently true
- 6:22
- And yet at the time
- 6:25
- our common sense was unique
- 6:29
- Nobody would have created the technology
- 6:31
- nobody would have created a company
- 6:32
- with the sole purpose
- 6:34
- of building technology to make video games possible
- 6:37
- And so that was our perspective
- 6:40
- Now, we felt that
- 6:41
- that, video games would, would of course fuel the technology
- 6:44
- development
- 6:46
- but you could use this technology for a
- 6:47
- whole bunch of other
- 6:48
- other reasons
- 6:49
- And, one of my favorite
- 6:51
- favorite, applications, this happened about, I guess about five years ago
- 6:58
- a small company, struggling company here in Silicon Valley
- 7:02
- called Keyhole
- 7:04
- And, they were, they created a, a 3D virtual world
- 7:09
- and it had no application
- 7:12
- This 3D virtual world
- 7:14
- and you start out in space
- 7:16
- you see the Earth
- 7:17
- you zoom into any location you wanted just
- 7:19
- by typing in the address
- 7:21
- I thought it was such a
- 7:22
- fabulous way of exploring the world
- 7:24
- going to places you've never been
- 7:27
- and they couldn't raise a penny
- 7:29
- And so, we, we, I was so excited about the company
- 7:33
- we put money into the company
- 7:35
- and I went everywhere and showed that demonstration
- 7:38
- I would tell people that this is the
- 7:39
- way we're going to do search someday
- 7:41
- If you want to search for something
- 7:43
- look for an address
- 7:44
- you would type it in
- 7:45
- and we would fly you there
- 7:47
- Okay?
- 7:48
- And, and, satellite images will continue to download
- 7:51
- and before you know it
- 7:52
- you're right there on the street
- 7:53
- and you might even see some buildings
- 7:55
- that, that, small company, was, eventually purchased by Google and
- 8:00
- became Google Earth
- 8:02
- Google Earth is now
- 8:04
- the single largest downloaded
- 8:06
- most
- 8:07
- frequently downloaded application in the history of mankind
- 8:11
- Over two hundred plus million downloads
- 8:15
- so 3D graphics could be used for a lot
- 8:18
- for a lot more than video games
- 8:19
- Now, that, that vision, if you will, that perspective
- 8:26
- was unique at the time
- 8:27
- and hard to sell
- 8:29
- And so
- 8:30
- we had to go and explain it to
- 8:31
- venture capitalists
- 8:33
- who
- 8:33
- had to figure out whether the technology was
- 8:35
- going to be possible
- 8:36
- how big was the market
- 8:37
- because it was zero billion dollars at the
- 8:39
- time
- 8:40
- So how do you extrapolate
- 8:43
- how do you
- 8:43
- scope the size of a market
- 8:46
- when its
- 8:47
- apparent size was zero at the time?
- 8:49
- And you look at
- 8:50
- analyst reports and, you study market research
- 8:53
- and all of it was sitting approximately zero
- 8:56
- It would never show up
- 8:57
- It's a non-category, a non-market
- 9:01
- and so
- 9:02
- it, you know, it's incumbent upon, the, the venture capitalist
- 9:06
- and of course the founders
- 9:08
- to try to figure out how to inspire
- 9:10
- each other into doing something together
- 9:12
- And so, Sequoia Capital and Sutter Hill were our venture capitalists
- 9:16
- and, we got the company going with two million dollars
- 9:21
- Now, the, the question about perspective
- 9:25
- becomes, very interesting, in other, in other opportunities in the company
- 9:30
- other, other circumstances in the company
- 9:33
- And let me give you some examples
- 9:35
- many years later, Sequoia Capital, came to me and said
- 9:40
- you know, there's a couple of kids at Stanford
- 9:44
- and they have this thing
- 9:47
- and it's a, it's an internet thing
- 9:50
- and you just type in
- 9:52
- you know, what you're looking for
- 9:54
- and it shows up
- 9:54
- you know, puts up the web
- 9:56
- website
- 9:57
- And
- 9:58
- I said, yeah, Yellow Pages, I mean, you know
- 10:01
- No duh, right?
- 10:03
- We use it, we, and, and there's a variety of
- 10:06
- of
- 10:06
- versions of it on the web at the
- 10:07
- time
- 10:09
- we use the internet just like everybody else to
- 10:11
- to do FTP and also to
- 10:13
- visit various websites
- 10:15
- and, and they said
- 10:16
- should we invest in this company?
- 10:17
- And I said
- 10:18
- there's no freaking way they're gonna make money
- 10:19
- That stuff is free
- 10:22
- right?
- 10:22
- And so they said
- 10:23
- well, you know, we can't figure it out ourselves
- 10:26
- but it doesn't cost much to give them a
- 10:27
- a million or two
- 10:29
- and, and they invested in a small company called
- 10:31
- what eventually became Jerry's company called Yahoo
- 10:35
- notice, although I had the perspective about one thing
- 10:39
- I didn't have the perspective about something else
- 10:41
- Just because you're a visionary doesn't mean you're
- 10:42
- a visionary about everything
- 10:44
- Your perspective stems from your life experiences
- 10:48
- what's commonsensical about you
- 10:50
- what's interesting to you
- 10:53
- and, and so that's important to
- 10:55
- to realize that you have perspective too
- 10:57
- Therefore, you have vision too
- 11:00
- Now, what's interesting about these websites
- 11:02
- like, to, to, to follow on the Yahoo story
- 11:05
- if you remember
- 11:05
- there were several other searches out there
- 11:09
- AltaVista at the time
- 11:13
- Excite at the time
- 11:15
- Lycos at the time
- 11:16
- right?
- 11:17
- And now the question is
- 11:18
- they're all doing search
- 11:19
- and they all did
- 11:20
- did a reasonably good job
- 11:23
- Now comes the question is
- 11:24
- what was their perspective?
- 11:27
- How did they
- 11:28
- how was their perspective different from one another?
- 11:30
- One website thought that they were a destination
- 11:33
- Do you guys remember that?
- 11:35
- We would be a destination
- 11:36
- kind of like a channel
- 11:38
- somebody said, in fact, since we're gonna be a destination
- 11:44
- we would serve up content
- 11:47
- And therefore
- 11:48
- the search part of it is a commodity
- 11:50
- We'll outsource that
- 11:52
- So all of the search engines
- 11:54
- which started out as search
- 11:56
- turned into destinations
- 11:58
- Or portals, and they outsourced the search to someone else
- 12:02
- which made it possible for Google to start
- 12:06
- And so notice two companies doing exactly the same thing
- 12:10
- started with the exact same fundamental core technology
- 12:18
- ended up in radically different places
- 12:20
- because they had different perspectives
- 12:23
- They saw the world differently
- 12:25
- So perspective matters, vision matters
- 12:28
- Now in our industry
- 12:30
- shortly after we were started
- 12:33
- 3D graphics for PCs and
- 12:35
- and consumer 3D graphics became the hottest
- 12:38
- hottest thing
- 12:39
- And so
- 12:39
- everybody in Silicon Valley was starting a 3D
- 12:41
- graphics company
- 12:43
- We were, in 1993
- 12:45
- the only consumer 3D graphics company in the world
- 12:48
- Silicon Graphics, up the street was the professional
- 12:52
- if you will, 3D graphics company
- 12:54
- By the end of a couple of years or so
- 12:56
- 1995, there were probably fifty
- 12:59
- seventy startups doing exactly the same thing we
- 13:01
- were trying to do
- 13:04
- And over time
- 13:05
- we competed with about two hundred companies
- 13:09
- NVIDIA today is the only surviving
- 13:12
- computer graphics company in the world
- 13:14
- And so the question is then
- 13:15
- what happened?
- 13:18
- Competition is intense
- 13:20
- Everybody has smart people
- 13:23
- Everybody has money
- 13:25
- We competed with IBM
- 13:27
- We competed with HP
- 13:28
- We competed with Silicon Graphics
- 13:32
- Competed with Sony, 3D effects, S3, Cirrus Logic, big, small, international, local
- 13:38
- We competed with companies all over the world
- 13:40
- So the question is
- 13:40
- what happened?
- 13:42
- I would argue that
- 13:44
- that, three hundred companies armed with exactly the same technology
- 13:50
- armed with exactly the same people
- 13:55
- The company that wins
- 13:56
- and let's say they all execute
- 13:58
- and they did, with three hundred companies
- 14:00
- you know
- 14:00
- fifty percent of them were going to execute
- 14:02
- at any given point in time
- 14:04
- And so the question is
- 14:05
- why does one survive?
- 14:07
- Well, I think that it matters to have perspective
- 14:09
- and let me give you some examples
- 14:12
- I always believed that
- 14:15
- you need to understand
- 14:17
- the reason why your business work
- 14:20
- What is the essence of your business?
- 14:22
- What makes it work?
- 14:24
- Now, the foundation of my business
- 14:26
- at its core is semiconductor technology
- 14:30
- Here in Silicon Valley
- 14:31
- we
- 14:32
- we usually like to refer to semiconductor technology
- 14:34
- as
- 14:35
- Moore's Law. Moore's Law is not so much
- 14:36
- a physical law as it's a law of competition
- 14:40
- it is a law of
- 14:42
- challenging engineers
- 14:44
- it's a law almost of setting pace
- 14:47
- And Moore's Law approximately
- 14:51
- Gives you twice the performance
- 14:52
- every year or two
- 14:55
- And so
- 14:58
- understanding the, the fundamental ingredient of our
- 15:01
- of our business improves by a factor of two every year
- 15:07
- and simultaneously
- 15:08
- reduces in cost by a factor of two
- 15:11
- every year
- 15:12
- The question is, what makes a survivable business?
- 15:15
- And so our first perspective
- 15:17
- was that 3D graphics was insatiable
- 15:22
- It was insatiable
- 15:24
- That if I made something twice as good every year
- 15:28
- even if the customer never asked for it
- 15:32
- even if the customer told us it was too expensive
- 15:37
- Even if the customer
- 15:38
- when you went to float that product specification to them
- 15:41
- told you that they're not interested
- 15:43
- and in fact, that was the case
- 15:45
- I took, I took our product spec to
- 15:47
- to Dell and HP and IBM and Gateway
- 15:50
- and they all told me it was too
- 15:51
- much money
- 15:52
- You're well outside of the boundaries of what
- 15:54
- they're willing to pay for
- 15:57
- When your customers all tell you not to do something
- 16:02
- the question is then
- 16:03
- what do you do?
- 16:04
- In our case
- 16:05
- because we had this unique perspective that 3D graphics was insatiable
- 16:11
- and Moore's Law was our friend
- 16:13
- therefore
- 16:14
- we should make our graphics processors twice as
- 16:17
- good every year
- 16:18
- And so for the first five years of our company
- 16:21
- we just turned off our blinders and said
- 16:23
- we're gonna ignore customers
- 16:26
- Now, which one of your
- 16:27
- of you guys are gonna go through your marketing
- 16:29
- marketing courses
- 16:30
- and the lesson that it teaches you is
- 16:32
- ignore your customers
- 16:35
- Well, sometimes you have to ignore your customers
- 16:37
- And the reason for that is because
- 16:38
- they don't know the nature of your business
- 16:41
- And while that industry is being created
- 16:45
- before there's common sense about the rules of that business
- 16:47
- there is no way they can possibly know
- 16:51
- And so we, I
- 16:53
- I took the last few million dollars of
- 16:54
- the company's money and
- 16:56
- built a chip that is way
- 16:57
- way, way too big
- 16:59
- And, our customers told us they
- 17:01
- we were way out of bounds on cost
- 17:03
- and they weren't going to buy any
- 17:04
- until the day we showed up with the
- 17:05
- processor
- 17:07
- We were in allocation through
- 17:09
- throughout the entire life of that project until
- 17:11
- our next generation product
- 17:13
- which was twice its price
- 17:14
- cannibalized the previous one
- 17:16
- And so we grew and grew and grew
- 17:19
- for several years
- 17:20
- Then the question became
- 17:22
- what now?
- 17:23
- Now
- 17:23
- you guys are going to learn
- 17:24
- that innovation is rather dangerous thing
- 17:28
- On the one hand
- 17:30
- once you discover a great idea
- 17:34
- Once you discover, discover a great idea
- 17:37
- you rinse and repeat
- 17:38
- rinse and repeat
- 17:39
- and you make that idea better and better
- 17:40
- and better
- 17:41
- Whether it's a laptop computer that you guys have here or
- 17:45
- a car or a microprocessor
- 17:46
- or in our case
- 17:47
- a graphics processor
- 17:49
- we made it better and better every year
- 17:52
- At some point, it becomes good enough
- 17:56
- Moore's Law is a wonderful thing
- 17:58
- Semiconductor technology
- 18:00
- enables you to make amazing leaps and bounds in technology
- 18:04
- and at some point
- 18:04
- it becomes good enough
- 18:06
- And so in the
- 18:08
- this is probably in the late nineties
- 18:12
- about seven years into our company
- 18:14
- maybe six, seven years into our company
- 18:17
- I came to the conclusion
- 18:21
- that 3D graphics was not going to be sustainable as a
- 18:28
- Accelerator
- 18:29
- or a fixed function device
- 18:31
- that renders texture maps and
- 18:34
- polygons on the screen
- 18:36
- That we had to change the company
- 18:40
- to make the 3D graphics processor programmable
- 18:44
- so that it could be
- 18:44
- an artistic medium for expression
- 18:47
- Now this is a weird word
- 18:49
- Here we are an engineering company
- 18:52
- and we now want to change this
- 18:55
- chip to become an artistic expression
- 19:00
- an artistic medium, so that all of the video games
- 19:03
- so that all of the applications that were
- 19:05
- developed on our chip would be stylistically different
- 19:11
- And we believed
- 19:12
- unless we could figure out a way
- 19:14
- to make the content richer and
- 19:17
- more interesting and stylistically different
- 19:20
- from one game developer to another game developer to
- 19:23
- you know, another, we would limit the life of our medium
- 19:28
- and if our medium reached this end
- 19:31
- we as the wor- the world leader
- 19:33
- would also, would also see our end
- 19:37
- And so we decided to make the GPU programmable and
- 19:42
- make it a medium for artistic expression
- 19:44
- and invented a technology called programmable shader
- 19:48
- Almost every single
- 19:49
- video game that you guys see today has
- 19:51
- our fingerprint on it
- 19:52
- whether it's a Xbox 360 or
- 19:55
- PS3 or any PC game today
- 19:58
- you could see elements of what
- 20:00
- programmable shaders made possible
- 20:04
- That started a whole new innovation curve for
- 20:06
- us and kept our industry vibrant
- 20:10
- but
- 20:10
- the crossing
- 20:12
- from one generation of technology to the next
- 20:14
- generation almost killed the company
- 20:16
- And so
- 20:18
- that process of reinventing the company
- 20:22
- The perspective that led us to a new
- 20:23
- idea also risks the company in the process
- 20:27
- and those are
- 20:28
- interesting conversations that we can have
- 20:30
- So there's perspective matters
- 20:32
- When a company gets larger
- 20:34
- you guys are, you guys are gonna learn that
- 20:36
- that, as the, the founder or as the CEO
- 20:40
- you have to learn new things
- 20:42
- And many of the new things that you'll
- 20:44
- learn has to do with
- 20:47
- building products at first
- 20:49
- and I've just talked to you about building
- 20:51
- products
- 20:52
- Soon you'll be talking about
- 20:54
- and learning about building companies
- 20:57
- And building companies means
- 20:59
- things that are soft and hard to explain
- 21:02
- like building a company with a culture
- 21:05
- What does that mean?
- 21:06
- How does the culture of one company different
- 21:09
- from a culture of another company?
- 21:10
- And why is it that this particular culture
- 21:12
- is better for your company and not for
- 21:14
- another?
- 21:15
- So the culture of a company is important to find out
- 21:18
- to, to, to
- 21:19
- to put your arms around and to create
- 21:21
- and develop
- 21:22
- how do you organize?
- 21:25
- Are you
- 21:25
- we were just talking earlier with one of the guys
- 21:27
- are you functionally organized? Are you
- 21:29
- organized in business units?
- 21:31
- How do you deal with multiple products and
- 21:32
- multiple geographies and multiple customers?
- 21:35
- And so that's the company building process
- 21:39
- It's mechanical, it's interesting, lots of trial and error
- 21:44
- it's organic, people matter, personalities matter
- 21:49
- and if you guys are interested in talking about that
- 21:51
- I'm happy to talk about that as well
- 21:53
- And then to, I would say probably the
- 21:56
- the, the most important thing
- 22:00
- above that is to realize that
- 22:04
- building a, when you're building a company and building a product
- 22:07
- skill matters, intellect matters, training matters
- 22:13
- But it's not enough
- 22:16
- The part of it that
- 22:16
- that
- 22:19
- that is important to
- 22:20
- to realize about building companies
- 22:23
- is that it's a challenging and painful and
- 22:28
- oftentimes, extraordinarily scary thing to do
- 22:33
- And so unless you have passion
- 22:35
- unless you really love the process of building
- 22:37
- the company and what you're trying to do
- 22:39
- it's going to be incredibly
- 22:40
- incredibly challenging
- 22:42
- And so what I would
- 22:44
- I would, I would, leave you with is
- 22:48
- when you're building a company
- 22:49
- if you decide to build a company
- 22:51
- you have to ask yourself
- 22:52
- what is the purpose that you're
- 22:55
- building the company for?
- 22:57
- Is it that you would like to build
- 22:58
- a company so that you can sell it
- 23:00
- make a fortune?
- 23:01
- is it that you would like to build
- 23:02
- a company so you can take it public?
- 23:05
- you're just a serial entrepreneur
- 23:07
- You want to build something
- 23:08
- let somebody else run it
- 23:09
- build something, sell it
- 23:12
- whatever your reason is
- 23:13
- be honest to yourself
- 23:15
- Turns out that, in my case
- 23:18
- I just love the process of building things
- 23:22
- and I love being part of something
- 23:26
- You know, being part of NVIDIA and being part of
- 23:29
- a, a people and being part of a cause that is
- 23:33
- that's inspiring to me
- 23:36
- keeps me vibrant and
- 23:37
- and it's something that I'm willing to do
- 23:38
- for a very long period of time
- 23:40
- And so I've been in
- 23:41
- in it now for sixteen years
- 23:44
- and
- 23:45
- I've learned a lot in the process of
- 23:46
- being the CEO and the founder of the
- 23:47
- company
- 23:48
- And so maybe the thing to do is
- 23:49
- why don't I open it up now and
- 23:52
- let you guys ask me
- 23:53
- whatever type of questions you guys have
- 23:56
- Yes, sir. What did your invent
- 23:58
- what did your investors say about
- 24:01
- your idea of, reinventing your company to include
- 24:04
- programmable chip, GPUs?
- 24:08
- yes, I will
- 24:09
- The question is, what did our investors
- 24:11
- say about, about us
- 24:15
- taking the big risk of adding programmable shaders and reinventing
- 24:21
- reinventing that product category and reinventing our company?
- 24:25
- first of all, it, it's not a
- 24:26
- it's not a conversation you really have with your
- 24:28
- with your shareholders, but you do have the conversation with your
- 24:33
- your management team first
- 24:36
- your employee second, your board of directors third
- 24:40
- typically is the way
- 24:41
- the process that I
- 24:43
- I take
- 24:45
- and
- 24:47
- when you're in, when you're in high tech
- 24:49
- when you're in a technology industry
- 24:50
- when the technology moves this fast
- 24:54
- if you're not reinventing yourself
- 24:56
- you're just slowly dying
- 25:00
- You're just slowly dying
- 25:01
- unfortunately, at, at the rate of Moore's Law
- 25:05
- Which is the fastest of any rate that we know
- 25:08
- right? The compound rate of Moore's law is
- 25:10
- pretty unbelievable
- 25:17
- I know, so it's scary
- 25:19
- And so you have to
- 25:21
- and this is, so the question is
- 25:23
- but we have a product that's generating a lot of money
- 25:25
- and it's very successful
- 25:26
- how do you cannibalize it?
- 25:29
- there is, there is a
- 25:30
- there is a theory that if you don't cannibalize it
- 25:32
- someone will, and it surely will be cannibalized
- 25:34
- And so, if you want to be a market leader
- 25:38
- you have to take the initiative to cannibalize
- 25:40
- your own products
- 25:41
- and have your ideas cannibalize your own ideas
- 25:45
- when we, when we, went from, a fixed function graphics accelerator
- 25:51
- a texture mapping engine
- 25:54
- for games like Quake
- 25:55
- III and Doom and those kind of games
- 25:58
- to a programmable shading architecture
- 26:01
- our first chip almost killed the company
- 26:04
- It was called GeForce FX
- 26:05
- I don't know if any of you have
- 26:07
- have ever, owned one of those
- 26:09
- GeForce FX is a
- 26:10
- is a chip, is a processor that
- 26:14
- you know
- 26:15
- it's a baby only a mother could love
- 26:17
- I mean, it's
- 26:21
- We, we, we took enormous chance in building GeForce FX
- 26:26
- but it almost broke our back
- 26:27
- But if it, if we didn't build that chip
- 26:31
- I am sure NVIDIA would be dead today
- 26:34
- I am absolutely certain we'd be dead
- 26:35
- It was one of the
- 26:36
- one of the biggest gambles in our history
- 26:39
- we had to create
- 26:41
- in, in, instead of, an API
- 26:44
- we had to go to a processor with a language
- 26:47
- we call it CG
- 26:48
- and a compiler, so it's kind of like a
- 26:50
- like a processor
- 26:52
- we introduced a new programming paradigm to the
- 26:54
- world that it never understood in the beginning
- 26:57
- And so it took a lot of evangelism
- 26:58
- a lot of marketing
- 27:00
- a lot of education
- 27:02
- but CG
- 27:04
- Inventing CG, took us to unbelievable places
- 27:08
- and, one of the, one of our most important
- 27:11
- work today, is related to using GPUs for general purpose computing
- 27:16
- and it's, extraordinary the, the results of what we're seeing
- 27:19
- and it wouldn't have been possible if it
- 27:20
- wasn't because of CG
- 27:22
- that we, we started then
- 27:24
- Okay, so you, you have to take these leaps
- 27:27
- Questions? Yes, sir
- 27:31
- some people are sort of see that gaming
- 27:33
- is sort of moving more towards consoles and
- 27:35
- away from PCs. Yeah
- 27:37
- Is it harder to make profit with
- 27:39
- you know
- 27:39
- when you have to negotiate with Sony or Nintendo or whoever
- 27:42
- some big company
- 27:43
- than it would be when you
- 27:45
- sell individual cards?
- 27:47
- Yeah
- 27:49
- the economics, the economics of it
- 27:52
- of
- 27:52
- building anything ultimately
- 27:55
- comes down to
- 27:57
- the amount of competition you have
- 28:00
- You don't set the price
- 28:01
- the competition sets the price
- 28:02
- The market doesn't set the price
- 28:03
- the competition does
- 28:05
- And so, so, if your competitor, wishes to build PlayStation Three
- 28:13
- as much as you do
- 28:15
- then, then the economics will be challenging
- 28:19
- And so, so
- 28:22
- there you just need to decide
- 28:23
- is it a
- 28:24
- is it an economic decision for you?
- 28:27
- for example, there's, there are rumors that
- 28:30
- that we were
- 28:31
- not enthusiastic in building some of the game consoles
- 28:34
- and we were more enthusiastic in building other
- 28:36
- game consoles
- 28:37
- it came down to this for me
- 28:39
- I think the
- 28:40
- you have to realize what is the finite resource
- 28:44
- what is the scarce resource that you have
- 28:47
- as a manager?
- 28:48
- The function of a manager
- 28:51
- is
- 28:52
- to allocate resources properly for the best return
- 28:57
- And so If you think about our resources
- 29:01
- our resource
- 29:02
- is the finite number of extraordinary engineers
- 29:07
- and how much time they have in a
- 29:09
- day to pursue whatever opportunities that are out there
- 29:13
- Now, if
- 29:16
- the number of opportunities that are out there
- 29:19
- is less than my supply of engineers
- 29:22
- so therefore demand exceeds is less than my supply
- 29:25
- then obviously I'm very enthusiastic about it
- 29:26
- But if it's the other way around
- 29:29
- then
- 29:29
- the opportunity to build a game console
- 29:32
- at terrible economics
- 29:34
- or any project at terrible economics
- 29:37
- is simply not worth it
- 29:39
- And so
- 29:39
- I look at it irrespective of competition
- 29:43
- The competition sets the price
- 29:44
- but then
- 29:45
- I get to decide whether I want to
- 29:46
- engage in that project or not
- 29:48
- You are in charge of your own company
- 29:50
- as the CEO
- 29:52
- And so we decide whether
- 29:54
- whether it's economic
- 29:55
- And, once you decide, then, you know
- 29:57
- it is what it is
- 29:58
- So you
- 29:58
- you have to be thoughtful about what is
- 30:00
- your critical resource?
- 30:01
- Do you have more of it or less
- 30:03
- of it than the market demands?
- 30:04
- Do you have more opportunity or less opportunity than the
- 30:06
- than, than your, than your resource can
- 30:09
- support?
- 30:09
- And then what's the appropriate return on that investment
- 30:12
- thinking about not just your cost
- 30:14
- but more importantly, your opportunity cost?
- 30:17
- and so
- 30:18
- so we look at it from that perspective
- 30:19
- every single time
- 30:22
- Can you tell us a little bit about
- 30:23
- the culture that you tried to set at NVIDIA?
- 30:25
- Yeah, that's a good question
- 30:28
- the question is, what is the
- 30:29
- can I talk about the culture that we're
- 30:31
- trying to set at NVIDIA?
- 30:33
- at the core of our company's
- 30:36
- success is innovation
- 30:39
- Now
- 30:39
- a lot of companies say innovation is
- 30:42
- important to their company
- 30:45
- invention's important to their company
- 30:48
- however, I don't believe you can fundamentally say that innovation
- 30:54
- that you want as a CEO
- 30:56
- to nurture the spirit of innovation
- 30:58
- to encourage innovation
- 31:01
- unless you have a culture of risk-taking
- 31:08
- We have to encourage our
- 31:10
- our engineers, excuse me, our marketing people
- 31:15
- all of our employees to take calculated risks
- 31:19
- In order to encourage them to take calculated risks
- 31:23
- first of all
- 31:23
- you have to teach them
- 31:24
- how to do that
- 31:25
- That's a skill, a matter of skill
- 31:27
- Then the second part of it is a
- 31:28
- matter of courage
- 31:31
- Most people hate to fail
- 31:35
- Are you guys, do you guys agree with that?
- 31:38
- Like, unless you guys want to be successful
- 31:43
- let me say it
- 31:44
- The positive way
- 31:45
- If you want to be successful
- 31:47
- I would encourage you
- 31:49
- to grow a tolerance for failure
- 31:53
- To develop a tolerance for failure
- 31:55
- Now, when I mean a tolerance for failure
- 31:57
- I don't mean, gee
- 31:58
- what Jensen just told me is sleep in until noon
- 32:02
- okay?
- 32:03
- Don't do any of my homework
- 32:05
- flunk out of all my classes
- 32:06
- because that defines failure
- 32:09
- right? That's not what I said
- 32:11
- What I said is
- 32:12
- what I'm trying to say
- 32:13
- is that I want you to try things
- 32:16
- even though it is impossible to calculate
- 32:20
- precisely that it would lead to success
- 32:22
- that your instincts
- 32:23
- your intuition is something you ought to follow
- 32:26
- If it wasn't because of
- 32:28
- following my
- 32:29
- own instincts
- 32:30
- or the founder's instincts or many of our employees' instincts
- 32:33
- why would we be where we are today
- 32:35
- and why would we have invented
- 32:36
- things that the markets never had before
- 32:38
- the world's never had before?
- 32:40
- So you have to have this
- 32:41
- this culture
- 32:43
- or tolerance for risk-taking
- 32:45
- But the thing about failure is this
- 32:48
- If you fail often enough
- 32:52
- you actually might become a failure
- 32:55
- And that's different than being successful
- 32:58
- And so the question is
- 32:59
- how do you teach someone how to fail
- 33:04
- but fail quickly
- 33:08
- and to change courses
- 33:09
- as soon as you know
- 33:10
- it's a dead end?
- 33:12
- And the way to do that
- 33:13
- is we call it intellectual honesty
- 33:18
- we assess
- 33:19
- on a per-
- 33:20
- on a continuous basis whether something makes sense
- 33:23
- or not
- 33:24
- And if it's the wrong decision
- 33:25
- let's change our mind
- 33:27
- And you know
- 33:28
- a lot of people say CEOs are always right
- 33:31
- and they never change their mind
- 33:33
- That doesn't make any sense at all to me
- 33:36
- especially when it violates the first principles
- 33:39
- of what we want the company to become
- 33:41
- An innovative company that invents amazing things
- 33:44
- that solves problems for the world that it
- 33:47
- sometimes didn't even know it had
- 33:51
- If you want to do that
- 33:53
- then you have to cultivate that
- 33:56
- tolerance for risk-taking
- 33:57
- and you have to then teach people how to fail
- 34:00
- but fail quickly and inexpensively
- 34:03
- and how to be direct with each other
- 34:06
- that
- 34:06
- that this is the wrong approach and what's the better approach
- 34:09
- and then, you know
- 34:10
- be flexible enough to change courses and quick
- 34:13
- And so that, that type of culture
- 34:14
- if you will
- 34:17
- in today's
- 34:19
- if you guys were to start a company
- 34:20
- and you were buil-building a
- 34:22
- a website with an internet service of some kind
- 34:24
- internet-based service of some kind
- 34:26
- with the competition coming from all over the world
- 34:29
- and it's 24/7, and ideas take no time to experiment
- 34:35
- And it, you know, a particular website or particular company
- 34:38
- company could be throwing ideas out into the world
- 34:42
- twenty a day
- 34:44
- And so unless you
- 34:47
- are thoughtful about risk-taking and being able to change your mind
- 34:51
- reacting to the market conditions
- 34:53
- and being flexible
- 34:54
- how are you going to stay alive?
- 34:56
- And so you could
- 34:57
- you could almost see
- 34:58
- what I just described
- 35:00
- in the nature of older companies
- 35:03
- and the nature of the newer companies
- 35:06
- The modern companies
- 35:08
- if you guys, you got
- 35:09
- I'm sure you guys all go to Google's website
- 35:11
- almost every single application's in beta form
- 35:16
- They're trying all kinds of stuff
- 35:17
- right?
- 35:18
- They're trying all kinds of stuff
- 35:19
- If they call it production and it doesn't work well
- 35:22
- you guys would just be upset at them
- 35:24
- So they call it beta
- 35:27
- Have you noticed? They call it beta
- 35:28
- so that they could try a lot of things
- 35:30
- and if it fails
- 35:31
- take it out
- 35:32
- If it's bad, take it out
- 35:33
- If it works, do more
- 35:34
- And so
- 35:35
- innovation requires a little bit of experimentation
- 35:38
- Experimentation requires exploration. Exploration will result in failure
- 35:42
- Unless you have a tolerance for failure
- 35:44
- you would never experiment. And if you don't ever experiment
- 35:46
- you would never innovate. If you don't innovate
- 35:48
- you don't succeed
- 35:50
- You'll just be a dweeb
- 35:54
- That's it
- 35:55
- Any other questions?
- 35:57
- Yes
- 35:57
- sir
- 35:58
- How did you choose your co-founders and initial team?
- 36:01
- don't ever, don't ever go into business with
- 36:05
- anyone you don't deeply trust
- 36:08
- And they were two of my
- 36:09
- two of my closest
- 36:11
- closest colleagues, and I trust them completely
- 36:14
- And, and, they're, they're wonderful friends even today
- 36:20
- So, by the way, as a CEO
- 36:22
- selecting people is ninety-nine percent of the job
- 36:26
- what applications do you see
- 36:28
- driving demand in the general purpose GPU market
- 36:32
- which is like
- 36:33
- what's going to be like the next big
- 36:35
- thing to send it to?
- 36:36
- So the question is
- 36:37
- what is
- 36:38
- what do I see driving the demand for this thing that
- 36:40
- that we're, we're pushing right now called
- 36:44
- GPU computing
- 36:45
- using the GPU for much more than just graphics
- 36:49
- for graphics, there's a, there's a, a model of graphics that
- 36:53
- we call computational graphics
- 36:55
- So, so it's
- 36:56
- using programs to generate the images
- 36:59
- the algorithms are no longer
- 37:02
- cast in the silicon
- 37:03
- the algorithms are actually software
- 37:05
- And it could be ambient occlusion
- 37:06
- it could be ray tracing
- 37:07
- it could be all kinds of interesting algorithms that
- 37:09
- that, people are
- 37:11
- are going to explore for the future
- 37:12
- We observed and it was
- 37:15
- it was in fact
- 37:17
- this is give you another example of innovation
- 37:19
- and, and, and for, with Stanford is, is fabulous
- 37:23
- we, when we invented GeForce effects
- 37:25
- although it wasn't a very successful GPU for graphics
- 37:29
- researchers around the world noticed that it had
- 37:31
- a programming language called CG
- 37:34
- C for graphics
- 37:35
- and that you could program this GPU to
- 37:37
- do other things aside from graphics
- 37:39
- And it had thirty-two bit floating point
- 37:41
- IEEE compatible thirty-two bit floating point
- 37:43
- And so some smart eng
- 37:46
- some smart researchers, many of them were here at Stanford
- 37:50
- just bought a graphics card from Fry's
- 37:52
- and started writing programs
- 37:55
- And they discovered that
- 37:56
- if they really worked hard and do all these men
- 37:59
- you know, these algorithmic gymnastics, they could get something
- 38:03
- an algorithm, you know, it could be nanomolecular dynamics
- 38:07
- could be computa, computational fluid dynamics
- 38:09
- to run twenty times faster
- 38:11
- And they couldn't believe it
- 38:13
- How do you speed up an application twenty times?
- 38:16
- Well
- 38:17
- the interesting observation that we made is that
- 38:19
- we speed up 3D graphics applications
- 38:22
- which is basically something you can do in software
- 38:24
- a thousand times over a CPU all the
- 38:26
- time
- 38:28
- So what if we took
- 38:29
- all of those parallel processors that were inside
- 38:31
- our GPUs
- 38:32
- and make it completely programmable and expose it
- 38:35
- through a programming language called C
- 38:37
- right? C and now
- 38:38
- and in your future C++
- 38:43
- imagine
- 38:44
- imagine the type of problems we could help
- 38:45
- solve
- 38:46
- And so whether it's weather prediction or seismic analysis or
- 38:51
- you know, taking your
- 38:52
- your CT scans and reconstructing the human image for the body
- 38:55
- the, the, the, the body from it
- 38:58
- all kinds of very computationally intensive applications
- 39:01
- we could accelerate fifty times
- 39:02
- a hundred times
- 39:04
- Now, just to put it in perspective
- 39:05
- a hundred times is a Moore's law time
- 39:08
- approximately ten years
- 39:11
- Now
- 39:11
- to put ten years' worth of computing resources
- 39:14
- in the hands of scientists
- 39:16
- researchers, engineers, unbelievable benefits
- 39:20
- And so, the risk was really large
- 39:22
- however, to make our GPU even more general purpose
- 39:26
- because every time you make something general purpose
- 39:30
- you know what, right, a Swiss Army knife?
- 39:32
- It's dangerous
- 39:33
- because
- 39:34
- whenever you make something general purpose or a Swiss Army knife-like
- 39:38
- you move away from your core business
- 39:43
- It's much, much better to have a very specific niche
- 39:46
- to have intense focus on a particular
- 39:49
- market segment. And you guys will learn all
- 39:50
- of this in marketing
- 39:53
- When you make something general purpose
- 39:54
- you're all things to all people
- 39:57
- you become, you know, what is it
- 40:01
- ma- what is it
- 40:02
- jack of all trades
- 40:03
- master of none
- 40:05
- Very, very dangerous move
- 40:07
- Now, we
- 40:07
- we thought that it was just too important
- 40:09
- for us not to do it
- 40:09
- so we decided to make that move
- 40:11
- And, it, it's, it's fabulous results
- 40:14
- Okay, so those are some of the
- 40:16
- some of the things that we're seeing now
- 40:18
- Yes, ma'am
- 40:32
- Okay, so her question is
- 40:34
- we were friends, in the beginning
- 40:36
- we're friends now
- 40:38
- How do we figure out
- 40:39
- who's the right position
- 40:41
- and how do we distribute the profits?
- 40:43
- Okay
- 40:45
- and so I, I won't say anything funny
- 40:48
- just so I'm not misunderstood
- 40:50
- but, but, all of our pay were identical
- 40:55
- and we all had identical share in the company
- 41:00
- so that's, that's, that's just simply fairness
- 41:05
- Now the question becomes governance
- 41:08
- There's the part of it which is equity
- 41:11
- Equity as another way of saying what's fair
- 41:14
- right?
- 41:14
- So we all had the same salary
- 41:16
- all
- 41:17
- all three of us were making a hundred
- 41:18
- thousand dollars a year
- 41:20
- okay? And we all had a percentage of the company
- 41:23
- equal percentage
- 41:25
- Now, you can't run a company
- 41:27
- though
- 41:27
- you can't build a great company
- 41:30
- when you have three people
- 41:31
- who has to vote on everything
- 41:33
- and with equal share of responsibilities
- 41:35
- You simply can't
- 41:36
- That becomes a leadership question
- 41:38
- that becomes a governance question
- 41:40
- that becomes a management question
- 41:42
- right? That becomes a question about
- 41:45
- building a great company
- 41:48
- I don't recall exactly
- 41:50
- exactly, the, the conversation
- 41:54
- but I think it kind of went like this
- 41:56
- All right, Jensen, you're the CEO, right?
- 41:58
- Okay
- 41:59
- That was done
- 42:02
- That was, that was basically the process
- 42:06
- I think that some people are
- 42:08
- are
- 42:09
- you know, I'm not particularly, I'm not
- 42:11
- but from a personality perspective
- 42:13
- I'm not particularly outgoing
- 42:15
- And so, that
- 42:16
- that's not a neces-necessity for being a good CEO
- 42:21
- but, as a personality
- 42:24
- I've always been able to see around the corners
- 42:27
- if you will
- 42:28
- I can see around the fuzzy edges
- 42:30
- And I think CEOs
- 42:33
- and, leaders need to be comfortable with ambiguity
- 42:39
- Ambiguity meaning that, you know
- 42:41
- what does the future look like?
- 42:43
- Well, it's hard to say
- 42:45
- Some people hate that
- 42:48
- Some people just say
- 42:49
- Jensen
- 42:50
- tell me what you need to have done and for how
- 42:53
- with how much resource and by when
- 42:55
- Okay?
- 42:56
- Some people rather, me tell them that
- 42:59
- hey, look, there's this, there's this opportunity out there
- 43:02
- not sure what it is
- 43:03
- not sure how big it is
- 43:05
- but it kind of feels like this
- 43:06
- let's go figure it out and let's build
- 43:07
- a business
- 43:08
- Some people can, are very comfortable with that
- 43:10
- And so this, this ambiguity, is, is important to, to, to, to be comfortable with
- 43:18
- I guess
- 43:18
- And I think that all CEOs that are
- 43:20
- are very successful are comfortable with ambiguity
- 43:23
- and I'm very
- 43:24
- I am very comfortable with ambiguity
- 43:27
- Yes?
- 43:28
- Yeah. What percentage of your initial investment was yours?
- 43:31
- Like, how do you get the rest?
- 43:34
- And also, like, how many times was your proposal for an
- 43:38
- for help with the investment was rejected?
- 43:42
- So I was thirty years old
- 43:43
- and I'd never taken
- 43:45
- A single business class
- 43:47
- and, and I've never taken any marketing classes
- 43:49
- and, and, and I've never used
- 43:52
- never used a, at the time, it wasn't PowerPoint, it was called Persuasion
- 43:57
- on the, on the Mac, it was called Persuasion
- 43:59
- And so, I bought a Mac so I could
- 44:03
- I could use Persuasion
- 44:05
- and, and, and then I tried to create a
- 44:09
- a company presentation
- 44:11
- to take it to venture capitalists
- 44:15
- The process kind of went like this
- 44:16
- We, started
- 44:17
- my first official day of work was my thirtieth birthday
- 44:20
- February seventeenth
- 44:21
- And, we got the company funded
- 44:25
- And so once we got started
- 44:27
- the question is, what are we going to do?
- 44:29
- You know, how does it all work out?
- 44:30
- How do we start the company?
- 44:32
- And so, we met every day
- 44:34
- the three of us
- 44:36
- in, in one of the founders' townhouse
- 44:38
- in Fremont
- 44:39
- And, and we would get together
- 44:41
- and, and there would be nothing to do
- 44:44
- I mean, what do you do?
- 44:45
- You got three guys that get together
- 44:46
- you just talk, you know?
- 44:48
- So, so what did you guys do last night?
- 44:50
- well, you know, what did you have for dinner?
- 44:51
- I mean
- 44:52
- so you talk about that for about six months
- 44:55
- Okay?
- 44:55
- And, and the big event of the day would be
- 44:58
- hey
- 44:58
- where do you guys want to go for lunch?
- 45:00
- And so, you know, Philly cheesesteak today, or, you know
- 45:03
- Chinese food tomorrow, or whatever it is
- 45:04
- That would be like a big deal
- 45:06
- And then after a while
- 45:07
- it was like
- 45:07
- could you put some donuts in the fridge
- 45:09
- in the morning for when we come?
- 45:10
- I mean
- 45:10
- so that would be a big deal for
- 45:12
- a while
- 45:12
- And so that lasted for a few months
- 45:14
- just the three of us like that
- 45:15
- I know it sounds pathetic
- 45:16
- but it's, it's, it's true
- 45:18
- Because at that time
- 45:20
- I'm reading about books on how to start companies
- 45:22
- and I'm trying to figure out
- 45:24
- you know, how to go raise money
- 45:26
- and, you know, what's a venture capitalist
- 45:28
- and how do you incorporate the company
- 45:30
- and so those kind of things
- 45:32
- pretty soon
- 45:33
- I met a, met a, met a lawyer, went to a
- 45:35
- went to a law firm called Cooley Godward
- 45:38
- and, they helped us incorporate the company
- 45:42
- and, the amount of money that he
- 45:44
- he, he says, you know, we need
- 45:46
- we need some money from you
- 45:48
- so that we could price the shares
- 45:50
- and also to incorporate the company
- 45:52
- So he says
- 45:52
- how much money do you have in your pocket?
- 45:54
- I said, two hundred dollars
- 45:55
- So he says, okay
- 45:56
- give me two hundred dollars
- 45:57
- I gave him two hundred dollars
- 45:58
- and, for two hundred dollars
- 46:01
- I bought, fifteen percent
- 46:03
- I think it was twenty percent of Nvidia
- 46:11
- So it was, it was a good deal
- 46:15
- Yeah, twenty percent, yeah
- 46:17
- And then I went
- 46:17
- I went back to the house and then
- 46:19
- went back to the condo and
- 46:20
- and, they all
- 46:21
- they both gave me two hundred dollars and
- 46:22
- they both got twenty percent
- 46:26
- And that's how it worked
- 46:27
- Literally
- 46:29
- It's not that much more
- 46:30
- You know, don't, here, here's, here's the thing
- 46:32
- Nvidia, I never finished my business plan
- 46:38
- I know it
- 46:40
- I know it
- 46:41
- we, we, we never finished, a business plan
- 46:46
- never could figure out how to finish a business plan
- 46:48
- to tell you the truth
- 46:50
- and, and if, if, if I would have finished that book
- 46:53
- and I, I went to, went to Borders and got
- 46:56
- Gordon Bell's book
- 46:58
- How to Start a High-Tech Company
- 47:00
- It's like this thick
- 47:03
- If I would have read the whole thing
- 47:04
- I would have been dead now
- 47:06
- We would have run out of money
- 47:07
- run out of time. And so I
- 47:08
- I, I read, I read the first three
- 47:10
- four chapters, and I, you know
- 47:12
- I, I, I got to go to work
- 47:13
- And so, so I, I
- 47:15
- incorporated the company, they introduced us to
- 47:18
- two venture capitalists
- 47:19
- and I just went to their office and
- 47:21
- told them what I like to do
- 47:23
- the, the thing that, that gets the company funded
- 47:26
- and when you're, when you get to that point
- 47:29
- you just have to remember a few things
- 47:32
- VCs don't invest in
- 47:35
- in business plans because business plans are easy
- 47:37
- to write
- 47:38
- I could write it
- 47:39
- but other people could
- 47:41
- right?
- 47:41
- And so, so, they invest in this
- 47:44
- They invest in great people
- 47:49
- And so, so the question is
- 47:51
- is, do they trust you?
- 47:54
- your reputation matters, your history matters
- 47:59
- because, because I had done so much work with
- 48:02
- Andy Bechtolsheim, which was another graduate of Stanford
- 48:05
- of Stanford, and the founder of Sun
- 48:08
- and the, and worked with
- 48:09
- the founders of Synopsys and LSI Logic
- 48:12
- and they, and, you know, we, we
- 48:14
- we were all very successful
- 48:16
- and we did good work
- 48:19
- Your reputation will precede you
- 48:22
- even if your business plan writing skills are
- 48:25
- are, you know
- 48:26
- inadequate. And the second thing is
- 48:28
- you need to have a vision that's sufficiently
- 48:30
- large to invest in
- 48:32
- because they, their statistics
- 48:34
- their probability of success is rather low
- 48:36
- And if they need to put in ten million dollars
- 48:39
- if the market's only twenty million dollars large
- 48:42
- they'll never get that ten million dollars back with
- 48:44
- with reasonable return
- 48:45
- But if it's a two hundred billion dollar market
- 48:47
- then of course it's a rather different thing
- 48:49
- Okay, so the size of the market
- 48:51
- And they want to know that
- 48:52
- that
- 48:53
- at least there is a clever idea
- 48:56
- that the market has never
- 48:58
- never done before. So they
- 48:59
- that last part is probably second
- 49:02
- you know, last, I said
- 49:03
- I said it last because also I think
- 49:05
- it's least important
- 49:06
- You have to
- 49:07
- you might have to reinvent yourself over time
- 49:09
- And if you want to reinvent yourself
- 49:11
- you need to have great people
- 49:12
- That's why great people is so important
- 49:15
- Yes, sir
- 49:18
- Who were some of the people that you
- 49:19
- considered to be your mentors when you were getting started
- 49:22
- and what was some of the best advice
- 49:23
- that you got from them?
- 49:25
- so the question is what
- 49:26
- what, who were some of the mentors and
- 49:29
- what were the best advice I got?
- 49:33
- I, I truly believe that if you want to be successful
- 49:36
- you, you
- 49:38
- a successful habit
- 49:40
- is to have the capacity
- 49:42
- and the willingness to learn from just about
- 49:43
- anybody
- 49:45
- And I do
- 49:46
- You know, I learn from just about anybody
- 49:49
- And it could be a little thing
- 49:50
- it could be a big thing
- 49:52
- you know, if it wasn't because of my kids
- 49:54
- I, I would be, I would
- 49:55
- I would, I would miss the whole internet age
- 49:59
- You know
- 49:59
- I would have missed YouTube and Facebook and Twitter and
- 50:03
- you know, I mean, without, so you, you need to, you need to know that
- 50:07
- that, the world changes and
- 50:09
- and you want to be able to learn
- 50:10
- from just about anybody
- 50:12
- And so I, I'm surrounded with extraordinarily talented executives and
- 50:16
- professionals, of, of all walks of life
- 50:19
- And so you just have to make sure that you're
- 50:21
- you're willing to learn from just about anybody
- 50:24
- some of the, some of the great advice
- 50:26
- that I've, I've had over the years
- 50:29
- focus, laser beam focus
- 50:32
- You know, don't do too much
- 50:33
- do, do a few things well
- 50:35
- and, do it with extraordinary intensity
- 50:38
- and, I, you know, focus, focus matters
- 50:42
- If you look at what I do with my time
- 50:45
- I wake up in the morning and the first thing in
- 50:46
- in my time is NVIDIA
- 50:47
- and the last thing I do is NVIDIA
- 50:48
- and I do that twenty-four seven
- 50:50
- And if I could figure out a way
- 50:51
- to do that for another fifty years
- 50:53
- we're going to be in good company
- 50:53
- Yes
- 50:56
- yes, ma'am
- 50:59
- Successful entrepreneur as well. So
- 51:10
- what's your greatest challenge as a president?
- 51:11
- Is your ambiguity as CEO
- 51:11
- what's your best and worst estimation of the
- 51:12
- future of you and your company?
- 51:13
- Okay
- 51:14
- So, her question is
- 51:19
- she started out by saying that
- 51:20
- that I'm also successful
- 51:21
- I'm a successful entrepreneur
- 51:23
- and, and, and what are my biggest challenges now?
- 51:27
- And, considering that I like ambiguity
- 51:30
- what's my best estimate of the future for the company or
- 51:33
- and, and for myself?
- 51:37
- the biggest challenge
- 51:38
- with building a company
- 51:40
- is the reinvention of the company
- 51:43
- Every successful thing
- 51:46
- needs to be torn down
- 51:48
- at some point and be rebuilt
- 51:52
- It is unfortunate but true
- 51:54
- And the reason for that is because the
- 51:55
- technology either gets good enough
- 51:59
- and therefore, you have to reinvent
- 52:01
- and sometimes the invention process is disruptive
- 52:06
- Sometimes it's in fact destructive
- 52:08
- and it could be
- 52:09
- it could destroy
- 52:11
- you could destroy what you have built in
- 52:12
- the past
- 52:13
- And so the reinvention process is very challenging
- 52:17
- it's gut-wrenching, takes a lot of courage
- 52:19
- and, it, it really tests your conviction
- 52:23
- In, in the technology industry
- 52:25
- reinventing the company
- 52:27
- every ten years is almost a necessary thing
- 52:30
- So that's, when I say challenging
- 52:33
- I don't mean bad challenging
- 52:34
- I think that's fun challenging
- 52:37
- I love the process of reinvention
- 52:40
- Okay, so that's fun challenging
- 52:42
- what's my best forecast for our company?
- 52:44
- I think that NVIDIA
- 52:45
- has the opportunity to become
- 52:47
- one of the most important technology company
- 52:50
- companies in the world
- 52:53
- And, I hope it
- 52:55
- I hope that it does
- 52:57
- And my best forecast for me is that
- 53:01
- I am eighty years old
- 53:03
- and I'm here talking to students
- 53:05
- and I hope I'm still the CEO
- 53:06
- Yes
- 53:11
- yes, sir
- 53:13
- questions about the first few years of when you start
- 53:16
- very critical, you have to be survive
- 53:18
- make sure your cash is positive
- 53:20
- your constant cash flow
- 53:22
- your cash is king
- 53:24
- So your Nvidia can talk to you when you get
- 53:26
- cash or, like
- 53:32
- yes, so the, the gentleman's question is
- 53:34
- has to do with
- 53:35
- in the beginning, survival's important, cash is king
- 53:40
- just so that there's no
- 53:41
- no, no ambiguity about this
- 53:44
- survival's always important, cash is always king
- 53:48
- And so as the CEO
- 53:49
- you're either making money
- 53:50
- saving money, or raising money
- 53:54
- And, and if you're not Making money
- 53:57
- raising money, and saving money
- 53:59
- you gotta be doing those three things
- 54:02
- It's a simple, it's a
- 54:04
- just stay focused on those three things
- 54:06
- And so when you're
- 54:07
- during the beginning, in the early days
- 54:09
- I was raising money all the time
- 54:11
- As soon as I was done raising this round of money
- 54:13
- I gotta raise more money
- 54:15
- You know, you're always raising money
- 54:16
- Just, maybe, maybe there was a week break in between
- 54:19
- but I was raising money all the time
- 54:21
- I was, as a startup, you're always going out of business
- 54:24
- right? That's the definition of a startup
- 54:26
- An enterprise that is nearly out of business
- 54:28
- all the time
- 54:32
- That's the definition of a startup
- 54:35
- Question?
- 54:36
- Yes
- 54:36
- ma'am. How are you prepared for the leadership
- 54:37
- succession for your company? Well
- 54:42
- because I wanted, I want this
- 54:43
- the question is how do I deal with leadership succession
- 54:47
- as a CEO and
- 54:47
- and for our company? Well
- 54:48
- I want this job until I'm 80
- 54:50
- I just said
- 54:51
- No, I'm just kidding
- 54:54
- one of the primary roles of a CEO
- 54:57
- in order to grow the company
- 55:00
- in order to grow the company
- 55:01
- in order to make
- 55:02
- make Nvidia one of the most important technology
- 55:04
- companies in the world and make
- 55:06
- make significant contribution to society
- 55:09
- in order to do that
- 55:11
- you have to cultivate new leaders
- 55:13
- so that they can have new ideas and
- 55:15
- grow new businesses and
- 55:17
- and, you know, maybe run a different geography
- 55:18
- run a new, new different product line
- 55:21
- And so I spend most of my time these days
- 55:24
- most of my time these days
- 55:26
- sitting with our general managers and sitting with
- 55:28
- our leaders and helping them think through
- 55:31
- strategies and helping them think through challenges and
- 55:33
- helping them think through
- 55:36
- product roadmaps and helping them think through transitions and
- 55:39
- you know, team building, organization creation
- 55:43
- you know, how to manage, how to
- 55:44
- how to create processes that
- 55:45
- that, last the test of time
- 55:47
- so, so, you know, these things are, are, lessons that
- 55:50
- that I'm supposed to pass on
- 55:52
- and I do
- 55:53
- and I spend a lot of my time
- 55:54
- doing that
- 55:55
- I believe, I believe this
- 55:58
- that succession planning by
- 56:02
- a priori picking out three people
- 56:06
- that the board should consider in the case
- 56:08
- that I get run over by a bus is a toxic
- 56:12
- toxic process
- 56:15
- I know that it has been
- 56:16
- it has been, thought of as a
- 56:19
- as a methodology
- 56:22
- a succession planning
- 56:23
- but I think it's just very toxic for
- 56:25
- the environment because
- 56:26
- everybody's trying to figure out who got selected
- 56:28
- and who didn't
- 56:30
- I think that it's a much
- 56:31
- much better process to focus on ultimately
- 56:34
- developing the next generation of leaders
- 56:37
- so that in the case that something happens
- 56:39
- where I'm not the right CEO anymore
- 56:41
- there are many choices for the board to choose from
- 56:44
- including outside
- 56:46
- So those are, I think the company building process
- 56:49
- focusing on that is the positive way of
- 56:52
- of thinking about succession planning
- 56:53
- A good question
- 56:56
- How about I take one question and that's it?
- 56:57
- Okay
- 56:58
- Yes, yes ma'am. It's okay
- 57:07
- Any question from
- 57:08
- from our Latino crowd here is fabulous
- 57:10
- I am Filipino. I have many friends who
- 57:10
- are engineers
- 57:14
- They have many, many brilliant ideas
- 57:18
- but they don't feel motivated because they
- 57:21
- they have money and they were looking for
- 57:24
- for people who could make
- 57:29
- Ideas
- 57:32
- real, and they are only working for other people
- 57:39
- and they don't realize these ideas
- 57:41
- I don't know what you see that is the problem
- 57:42
- Well, first of all, so, so her, her, I think
- 57:43
- let me see if I can paraphrase the question
- 57:45
- you, you know
- 57:46
- you know friends who have a lot of ideas
- 57:48
- but they also have a lot of money
- 57:49
- and so they don't really feel that motivated
- 57:51
- to do something with those ideas
- 57:54
- you know, you know, first of all
- 57:57
- money, I hope that if I
- 57:59
- if I leave you with anything
- 58:01
- money is the only singular reason
- 58:03
- not to start the company
- 58:06
- Because starting companies are
- 58:08
- it's a very, very unlikely
- 58:12
- probability for success
- 58:15
- And so if you're looking for
- 58:17
- if that is your reason for doing it
- 58:19
- you will likely regret the experience
- 58:22
- You
- 58:23
- you should start the company and build a
- 58:24
- company because you just
- 58:26
- you so believe in your idea
- 58:27
- you so, you're so passionate about it
- 58:30
- and you want to build something great
- 58:33
- The only reason you want to do it
- 58:35
- It just comes along
- 58:38
- I, I have, I have, I have plenty of money
- 58:42
- and, and, and, and, but, but it doesn't motivate me
- 58:48
- nor does it demotivate me
- 58:49
- because it, it wasn't
- 58:53
- And so, I would urge you to ask yourself
- 58:57
- and for your friends to ask
- 58:58
- ask themselves
- 58:59
- what is the purpose that they want to
- 59:00
- start a company?
- 59:02
- ideas?
- 59:04
- Ideas are a dime a dozen
- 59:06
- There's so many ideas
- 59:07
- If you want ideas
- 59:08
- you're going to get a lot of ideas
- 59:09
- in this room
- 59:11
- And so ideas don't really matter
- 59:13
- you have to have a
- 59:14
- a perspective that's unique
- 59:16
- that you feel really strongly about it
- 59:19
- and that you're willing to
- 59:20
- persevere almost any challenge
- 59:22
- That's the reason why you should start your
- 59:25
- career
- 59:26
- Okay?
- 59:27
- Thank you, everybody
Watch these videos on your TV
XTTV brings video content from X to your living room. Download now and start watching on the big screen.