
Nvidia CEO Jensen Huang Presents Without Rehearsing. Should You?
SAN JOSE, CALIFORNIA - MARCH 18: Nvidia CEO Jensen Huang delivers the keynote address during the ... More Nvidia GTC 2025 at SAP Center on March 18, 2025 in San Jose, California. (Photo by)
Nvidia CEO Jensen Huang delivered a much-anticipated keynote at GTC 2025, the company's annual developers conference, and did so 'without a net.' Huang's bold statement set the tone for a brilliant presentation which he famously does not rehearse.
'There are no scripts, there's no teleprompter,' Huang told the 25,000 attendees who had gathered at a San Jose stadium.
For the next two hours Huang unveiled the next generation of groundbreaking chips that will power artificial intelligence.
The New York Times dubbed Nvidia's conference 'The Super Bowl of A.I.' But Huang, unlike athletes who practice relentlessly for the big game, chooses not to rehearse for the spotlight. Instead, he prefers a 'grip it and rip it' style of presenting.
If you watch the presentation, even part of it, you'd think Huang had practiced for weeks. Huang delivered the presentation, demos, slides, and messages as effortlessly as someone who's rehearsed for a TED Talk hundreds of times—and Huang covered a lot more ground than speakers do in an 18-minute TED Talk.
Yes, Huang's approach looks authentic, casual, and conversational—a difficult feat to achieve in any professional business presentation, let alone a presentation that takes a deep dive into extraordinarily complex technology. Huang easily shifts between explaining complex technology and interacting with AI-powered robots on stage.
But just because Huang doesn't rehearse, should you follow his lead?
Don't even think about it.
Although Huang may not 'rehearse' in the traditional sense of the word, his keynotes are not as 'unscripted' as they appear. And that's because Huang knows the material better than anyone.
Before you present 'without a net,' remember that Huang has been leading the company he co-founded for over thirty years, making him one of the longest-serving CEOs in America. And it's nearly impossible to top his work ethic. Huang is famous in Silicon Valley for keeping a grueling schedule of fourteen-hour days, seven days a week.
In short, he knows his stuff. And he knows what's coming next in the presentation. According to Nvidia's marketing chief, Huang worked with product divisions for months before the conference and collaborated with marketing on slides and demonstrations.
Huang doesn't need a teleprompter because his mind is the teleprompter. Huang can speak for hours about any one of the technologies (or slides) the company makes. The only thing he needs to pay attention to is the time, so he doesn't run into the other scheduled events.
Huang's deep knowledge of the material allows him to speak 'off the cuff' about the history of technology cycles, the power of new chips, and the future of artificial intelligence.
'This slide is genuinely my favorite,' Huang said about 23 minutes into the keynote. The slide represented the company's core strategy and long-term vision, which Huang had been preaching for twenty years since Nvidia's first developers conference. In other words, he's practiced it for twenty years.
So, the next time you present a newly created slide that you've rehearsed twice, remember that Huang has been 'presenting' his vision for decades.
Most speakers shouldn't attempt Huang's 'grip it and rip it' style because if they don't share Huang's deep subject matter expertise and comfort with the message, the result will be a rambling, convoluted, and confusing presentation.
However, with a little practice, you can make even the most complex information engaging and conversational. According to research studies, practicing a presentation until you've internalized every slide will alleviate anxiety. If you feel confident about the content, you'll start enjoying the process instead of dreading it. Your audience will notice, too.
Rehearsing your presentation again and again—I recommend at least ten times before the actual event— can lead to several positive outcomes, in addition to building confidence and reducing anxiety.
Your pacing will become more natural, allowing you to speak conversationally rather than trying to memorize your lines.
The last point is critical. Using too many filler words is a sure sign a speaker failed to practice. For example, Jensen Huang's keynote was remarkably clear of fillers. I put the transcript of the keynote into a tool that analyzes grammar, and it returned just ten clear instances of 'um' or 'ah' in the entire 2-hour, 11-minute presentation. For context, one study found that college professors use about 3 to 5 filler words per minute in a typical lecture. That would translate into nearly 400 fillers in Huang's keynote.
Filler words are not inherently bad, of course. They're part of our natural conversation. But studies have found that too many fillers disrupt speech and leave an unfavorable impression on the audience.
Jensen Huang's unique presentation style works for him and it's highly engaging, but it's unlikely to work for you--unless you have thirty years of product knowledge. But with just enough practice you, too, can appear authentic, knowledgeable, and engaging in your next presentation.
Hashtags

Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Yahoo
4 hours ago
- Yahoo
Nvidia's Jen-Hsun Huang says Chinese competitors are 'quite formidable' just days after the announcement of a Chinese RTX 4060-level GPU powering up
When you buy through links on our articles, Future and its syndication partners may earn a commission. US export controls have increasingly stopped GPUs with certain AI capabilities from getting into China, and just this week, Nvidia stated the US's ban of H20 chips into the country meant a 'multibillion-dollar write-off' for the tech company. Though these protectionist policies from the US are not keeping China from building its own GPUs, as a new China-made GPU has powered up. Lisuan Technology, a Chinese startup, has been developing the "first self-developed architecture and fully independent intellectual property GPU chip", and it turned on this week. This was announced in a recent WeChat post, alongside additional guarantees to "carry out detailed and comprehensive software and hardware testing and driver optimization work." According to Tom's Hardware, the 6 nm GPU chip is targeting RTX 4060-level performance, and is currently titled the G100. Lisuan has reportedly been working on the G100 since 2023, with plans to launch it in 2023, so there is a level of skepticism around whether or not it can actually hit that RTX 4060 performance level. Despite being a budget card from the last generation, the RTX 4060 is still an impressive card, built on the 5 nm process from TSMC. The smaller the process, the higher the density of transistors, and this results in better performance and efficiency. Effectively, it's harder (and sometimes impossible) to get the same performance out of older processes. The G100 powering on is a good sign for the card, but it's the first of many steps before it can actually see a launch into the Chinese market. Further optimising and just plain testing is needed, especially as it's targeting "the needs of desktops, notebooks, graphic workstations and other devices." Huang tells Bloomberg, "The Chinese competitors have evolved" and Huawei, with its new AI chips, has become "quite formidable." Given that the domestic ability to create processes for the G100 is quite limited, it is likely that Huawei's Ascend 920 and Lisuan's G100 are using silicon from the same Chinese foundry: SMIC. This is all according to estimations from Tom's Hardware. According to Huang, 'Like everybody else, they [Chinese companies] are doubling, quadrupling capabilities every year." This is all to build towards a central point that Huang wants "all of the world's AI researchers and all of the world's developers to be building on American stacks". This is reportedly "irrespective of the near-term revenue success that we have", though opening up sales to a bigger market surely can't hurt the hardware giant. Best CPU for gaming: Top chips from Intel and gaming motherboard: The right graphics card: Your perfect pixel-pusher SSD for gaming: Get into the game first.
Yahoo
4 hours ago
- Yahoo
Nvidia's Hopper GPUs are now dead to the Chinese market after export controls that made the company take a 'multibillion-dollar write-off'
When you buy through links on our articles, Future and its syndication partners may earn a commission. There was more than the usual swell of anticipation for Nvidia's latest earnings call, primarily because the last quarter has been tumultuous in the wake of US tariffs and trade restrictions. On this front, and despite the fact that the AI chip giant still seems to be doing phenomenally well, Nvidia has admitted export controls have fully killed off its Hopper generation GPUs in China. During the company's recent Q1 earnings call, Nvidia CEO Jensen Huang explained: "The H20 export ban ended our Hopper Data Center business in China. We cannot reduce Hopper further to comply. As a result, we are taking a multibillion-dollar write-off on inventory that cannot be sold or repurposed. We are exploring limited ways to compete, but Hopper is no longer an option." Hopper is the company's previous-gen GPU/AI accelerator architecture. While its Blackwell architecture—the architecture at the heart of the RTX 50 series—is rolling out to fill up data centres despite previous delays, Hopper chips still line many server racks and they were the primary Nvidia export to China. The past couple of years have seen the same scene play out over and over again: The US restricts what Nvidia can export to China, Nvidia starts exporting a slightly less powerful Hopper chip to China, then the US restricts it further so that less powerful Hopper chip is restricted, too. Rinse and repeat. No longer, though, according to Nvidia. Now, there is seemingly no less powerful chip that Nvidia can comfortably make and export to the country. Nvidia Hopper is dead in China. Nvidia CFO Colette Kress says: "our outlook reflects a loss in H20 revenue of approximately $8 billion for the second quarter." H20 is the Hopper chip that Nvidia was previously exporting to China, and $8 billion revenue loss for Q2 is a lot more than the company lost for Q1. Nvidia had previously said that it could lose $5.5 billion in Q1 because of export restrictions, but it looks like that amount turned out to be $2.5 billion in the end: "We recognized $4.6 billion H20 in Q1. We were unable to ship $2.5 billion, so the total for Q1 should have been $7 billion." Despite praising President Trump's "bold vision", the company doesn't seem to agree with his trade restriction strategy in this case. Huang says: "The question is not whether China will have AI, it already does. The question is whether one of the world's largest AI markets will run on American platforms. Shielding Chinese chipmakers from U.S. competition only strengthens them abroad and weakens America's position." Your next upgrade Best CPU for gaming: The top chips from Intel and gaming motherboard: The right graphics card: Your perfect pixel-pusher SSD for gaming: Get into the game ahead of the rest. We've heard Huang say similar before, and it's certainly an argument to take seriously. At the same time, though, we can hardly expect the CEO of a chip company to support the banning of its exports to one of its biggest markets. The China export restrictions were certainly the main talking point in the earnings call, other than the usual "AI factory" stuff and a sliver of gaming talk. On that front, Nvidia claims a "record $3.8 billion" gaming revenue, but the wow-factor shrivels a little when we remember that Nvidia's pushed out a bunch of its new GPUs over a very short period, so we can expect an inflated number there. Nvidia all but admits this when it calls Blackwell its "fastest ramp ever"—that's "fastest", not "biggest". Anyway, trade talk aside, Nvidia seems to be doing pretty well in the wake of this news. I'm sure the multi-billion company will survive Hopper waving farewell to China. Error in retrieving data Sign in to access your portfolio Error in retrieving data Error in retrieving data Error in retrieving data Error in retrieving data
Yahoo
4 hours ago
- Yahoo
Nvidia's research boss claims the company's Chinese AI researchers are now writing programs for Huawei instead and is blaming the US chip exports
When you buy through links on our articles, Future and its syndication partners may earn a commission. Nvidia's been banging the drum against the United State's China chip export restrictions for a while now, but while it had previously highlighted this in broad terms, the company now seems to be getting more direct with its claims. According to a machine translation of a report from Taiwan Economic Daily (via Wccftech), Nvidia's chief scientist and senior VP of research, Bill Dally, claims that Huawei is scooping up ex-Nvidia AI researchers as a result of the restrictions. According to Dally, admittedly via a machine translation, the growth in the number of AI researchers working in China—apparently growing from a third of the world's researchers in 2019 to almost half of them today—has been forced by the US export restrictions. The idea is that without these restrictions, Huawei wouldn't be forced to lean so strongly into home-grown AI solutions, but now it must do so to keep up. Nvidia is clearly keen on presenting this argument (probably in hopes that the US administration specifically will hear it) to show that there are arguable downsides of banning its exports to China for the US. It certainly appeals to the ears of those concerned about the US-China technological arms race. As I said, though, the general argument isn't new—Nvidia has been touting it for a while. At Computex last month, Nvidia CEO Jensen Huang said: "AI researchers are still doing AI research in China" and "if they don't have enough Nvidia, they will use their own [chips]." And regarding Huawei specifically, Huang said the company has become "quite formidable". There is, of course, another reason other than US national interest that might make Nvidia keen to highlight possible negatives of export controls. Namely, the fact that these restrictions have cost and will cost the company lots of money. Nvidia itself has confirmed this, stating that after billions of dollars lost through restrictions of its H20 chips to China in Q1, it's expecting another $8 billion to be lost for the same reason in Q2. That's because Hopper, the company's previous chip architecture, "is no longer an option", according to the CEO. Huawei's latest Ascend 910 and 920 chips, courtesy of China's SMIC (Semiconductor Manufacturing International Corporation), will probably now be better options for Chinese AI companies than trying to get hands on Nvidia silicon somehow. And with ex-Nvidia researchers now apparently padding out the Chinese industry, who knows what will be cooked up next and when. Nvidia certainly seems to be presenting itself as worried about what's to come. The company can't complain about the vaguely 'poachy' aspect of this, though, really—not when Nvidia seems to be enticing likely TSMC employees in Taiwan with high salary job advertisements. Sometimes business is just business, you know? Best gaming PC: The top pre-built gaming laptop: Great devices for mobile gaming. Melden Sie sich an, um Ihr Portfolio aufzurufen.