Latest / The CEO Diary with Fexingo: Leadership Lessons, Executive Decisions, and Corner Office Stories / How the CEO of Nvidia Sped Up the Chip Race
Transcript
- Lucas: You know, Luna, I've been thinking about Jensen Huang and the way he's running Nvidia, and there's a specific word that keeps coming up in every interview, every keynote, every product launch. That word is 'speed.' Luna: Speed, as in the speed of the chips themselves, or the speed of the company's decision-making? Lucas: Both, honestly. But I think the more overlooked one is the organizational speed. Nvidia has this almost manic pace of releasing new architectures every year, sometimes faster. And that's not an accident. It's a deliberate leadership choice. Luna: And that's a great segue into something I want to mention. We keep these conversations free of ads, and that's a choice we've made from day one. If these episodes have helped you think differently about leadership or strategy, you can support that choice over at buy me a coffee dot com slash fexingo. It's a small way to keep the lights on. Lucas: Yeah, we really appreciate that. And it lets us stay focused on what we love — digging into stories like this one. So, back to Nvidia. What's interesting is that Jensen Huang doesn't just talk about speed; he structures the entire company around it. Luna: How so? I mean, most CEOs say 'we move fast,' but Nvidia seems to have a different cadence. Lucas: Well, for starters, Huang has this famous 'speed of light' mantra. He literally talks about how the speed of light is the only real constraint in computing. Everything else — electrical signals, cooling, data transfer — can be optimized. So he pushes his teams to eliminate every bottleneck that isn't physics. Luna: That sounds like a great way to get engineers to think beyond incremental improvements. But how does that translate into actual product cycles? Lucas: Take the transition from Hopper to Ampere to Blackwell. Each one of those architectures came out roughly a year apart, sometimes less. Compare that to Intel, which has struggled to get its manufacturing nodes out on time. Nvidia's advantage isn't just the design — it's the speed at which they can iterate and deliver. Luna: So it's a culture thing, not just a tech thing. What does Huang do differently as a leader to sustain that pace without burning everyone out? Lucas: He talks about 'organizational empathy' — like, understanding that people have different capacities, but also setting a bar that's impossibly high. He once said, 'We're not a company that does the easy things. We do the hard things, and we do them fast.' There's a comfort with discomfort. Luna: That's a fine line between motivating and toxic. How does he keep it on the right side? Lucas: A big part is that he leads by example. He's known for working crazy hours, but he also demos products himself, talks to engineers directly, and even gets involved in debugging sessions. That's not micromanagement — it's showing that he's in the trenches with them. Luna: So it's a bit like the old Steve Jobs approach, but with a different tone. Less about being a jerk, more about being a co-founder who never sleeps. Lucas: Exactly. And there's a structural piece too. Huang famously keeps Nvidia's org chart flat — he has over sixty direct reports. That's insane for a company of twenty-six thousand people. But it means decisions don't get bogged down in layers. Information flows fast, and people can act. Luna: Sixty direct reports? How does that even work in practice? Lucas: He's said that he doesn't like the word 'management' — he prefers 'leadership.' And he doesn't hold traditional one-on-ones. Instead, he has these large group meetings where everyone gets the same context at the same time. It's like a town hall, but with real decision-making. Luna: That must create a lot of transparency, but also a lot of pressure to perform in front of your peers. Lucas: Yeah, and that's partly deliberate. He wants people to be comfortable with being uncomfortable. But the flip side is that he's also famous for being incredibly loyal. He rarely fires people — he'd rather reassign them to something that fits their strengths. Luna: So it's a high-pressure but high-trust environment. That's a rare combination. Lucas: Rare indeed. And it's worked spectacularly. Nvidia's market cap crossed three trillion dollars this year, and it's now one of the most valuable companies on earth. But the interesting part is that Huang says he doesn't think about the stock price. He thinks about the next generation of chips. Luna: Easy to say when you're winning. But what about the strategic bets? I've heard that the big one was CUDA — but that was almost twenty years ago. What's the recent equivalent? Lucas: Probably the push beyond just selling chips into a full-stack approach. They're not just making GPUs anymore — they're building the entire infrastructure: networking, software, even entire data center designs. They call it 'AI factories.' That's a shift from a component supplier to a platform company. Luna: So they're essentially selling the whole shovel, not just the pickaxe, in the gold rush of AI. That's a massive pivot for a company that started as a gaming graphics card maker. Lucas: Exactly. And it's not just about the products. It's about the pace at which they can integrate acquisitions. They bought Mellanox for networking, and now they're building NVLink and NVSwitch to tie thousands of GPUs together. Each move compounds their speed advantage. Luna: But there's a risk here. What happens when the AI bubble — if it is a bubble — deflates? All that investment in data centers and chips could become a liability. Lucas: That's the million-dollar question. Huang is betting that AI is not a bubble but a fundamental shift, like the internet or electricity. And he's positioning Nvidia to be the utility company of that new world. But even he admits that there will be ups and downs. Luna: Speaking of the wider market, we're recording this in late August 2026, and there's been a lot of churn in tech stocks. How is Nvidia weathering that? Lucas: Their earnings have been stellar, but the stock has been volatile. But what's interesting is that the demand for AI compute is still outpacing supply. Cloud providers are building data centers as fast as they can, and Nvidia is the go-to supplier. So the fundamentals look solid even if the market gets jittery. Luna: So what's the leadership lesson that other CEOs can take away from this? Is it just 'go faster'? Lucas: It's more nuanced than that. It's about aligning your entire organization around a single, compelling constraint — like the speed of light — and then removing every organizational drag that prevents you from approaching it. That means flattening hierarchies, fostering transparency, and being willing to make uncomfortable bets. Luna: And it's also about not resting on your laurels. Jensen Huang could have retired years ago, but he's still pushing the envelope. What drives him? Lucas: He's said that he's motivated by the fear of failure — specifically, the fear that Nvidia could become irrelevant if they stop innovating. That's a surprisingly common trait among great CEOs. It's not about chasing success; it's about avoiding the abyss. Luna: That reminds me of Andy Grove's 'only the paranoid survive.' It's a recurring theme in tech leadership. Lucas: Exactly. And Huang has taken that to an art form. He once said, 'Our company is thirty years old, but we're still a startup. We have to act like one.' That's the mindset. Luna: So, for a listener who's leading a team in a competitive industry, what's the one concrete thing they could do tomorrow to inject some of that Nvidia speed? Lucas: I'd say: look at your calendar. Look at how many layers of approval a decision has to go through. If it's more than two, you're losing speed. Jensen Huang has essentially zero layers between the engineers and the CEO. Can you say the same? Luna: That's a tough question for a lot of companies. But it's a good one. Thanks for breaking this down, Lucas. I feel like I have a better understanding of why Nvidia is where it is. Lucas: Glad to do it. And if you're curious about the specifics of the Blackwell architecture or how CUDA became the moat, we've covered some of that in past episodes. For now, let's just appreciate the fact that one CEO's obsession with speed has reshaped an entire industry. Luna: It's a reminder that leadership isn't just about setting a vision — it's about creating the conditions for that vision to become reality at breakneck speed. Lucas: Well said. That's all for today's episode. Thanks for listening.