Navigate Select ESC Close

Jensen Huang – Will Nvidia’s moat persist?

2026-04-15 Science & Technology
153.3k
4.4k
1.0k
Dwarkesh Patel
Dwarkesh Patel
1.3m subscribers

Unlock all features

FREE: Get instant access to 10 AI summaries, chats, or transcripts per day.

Description

I asked Jensen about TPU competition, Nvidia’s lock on the ever more bottlenecked supply chain needed to make advanced chips, whether we should be selling AI chips to China, why Nvidia doesn’t just become a hyperscaler, how it makes its investments, and much more. Enjoy! +𝐄𝐏𝐈𝐒𝐎𝐃𝐄 𝐋𝐈𝐍𝐊𝐒 * Transcript: https://www.dwarkesh.com/p/jensen-huang * Apple Podcasts: https://podcasts.apple.com/us/podcast/jensen-huang-tpu-competition-why-we-should-sell-chips/id1516093381?i=1000761582962 * Spotify: https://open.spotify.com/episode/1viBRy6dQdlSw0OdFvogXB?si=bc2cdbd467ed4ee3 𝐒𝐏𝐎𝐍𝐒𝐎𝐑𝐒 - Crusoe's cloud runs on state-of-the-art Blackwell GPUs, with Vera Rubin deployment scheduled for later this year. But hardware is only part of the story—for inference, Crusoe's MemoryAlloy tech implements a cluster-wide KV cache, delivering up to 10x faster TTFT and 5x better throughput than vLLM. Learn more at https://crusoe.ai/dwarkesh - Cursor helped me build an AI co-researcher over the course of a weekend. Now I have an AI agent that I can collaborate with in Google Docs via inline comment threads! And while other agentic coding tools feel like a total black-box, Cursor let me stay on top of the full implementation. You can try my co-researcher out at https://github.com/dwarkeshsp/ai_coworker, or get started on your own Cursor project today at https://cursor.com/dwarkesh - Jane Street spent ~20,000 GPU hours training backdoors into 3 different language models, then challenged my audience to find the triggers. They received some clever solutions—like comparing the base and fine-tuned versions and extrapolating any differences to reveal the hidden backdoor—but no one was able to solve all 3. So if open problems like this excite you, Jane Street is hiring. Learn more at https://janestreet.com/dwarkesh To sponsor a future episode, visit https://dwarkesh.com/advertise. 𝐓𝐈𝐌𝐄𝐒𝐓𝐀𝐌𝐏𝐒 00:00:00 – Is Nvidia’s biggest moat its grip on scarce supply chains? 00:16:25 – Will TPUs break Nvidia’s hold on AI compute? 00:41:06 – Why doesn’t Nvidia become a hyperscaler? 00:57:36 – Should we be selling AI chips to China? 01:35:06 – Why doesn’t Nvidia make multiple different chip architectures?

Top Comments (10)

@bjarke7886 2026-04-15

I really appreciate how Dwarkesh always starts straight to the point. He isn't confrontational, but he asks questions that prompt real responses rather than prepared ones.

962 37 replies
@Rathgrams2 2026-04-15

You don’t have to move on. I’m enjoying it. -Jensen Huang

907 44 replies
@prajwalshetty6588 2026-04-15

love the heated interview, props to dwarkesh for tough interview

870 24 replies
@illegalplants 2026-04-15

so refreshing to see someone press Jensen, everyone else does sycophantic interviews, great work Dwarkesh!

801 16 replies
@bjrnfruderman13 2026-04-15

really appreciate you pushing harder questions than most interviewers. keep it up, even though its difficult at times.

317 6 replies
@IfThisDoesntWork 2026-04-15

My man said “I am the evidence.” It had a real “I am the one who knocks” vibe. 🔥

264 9 replies
@wonilart8426 2026-04-16

Jensen's argument has historical basis. The GAO found in 2002 that restricting semiconductor equipment exports to China was ineffective because Japanese and European firms kept selling, only ~2 generations behind state-of-the-art. Same story with satellites post-ITAR (Europe built an "ITAR-free" industry that now outcompetes us on price), machine tools (US makers wiped out, we now import from China), and drones (Turkey's Bayraktar, DJI's dominance). There's a historical pattern of unilateral controls accelerating indigenization in the target country while handing market share to allied competitors who aren't restricted. The revenue loss then starves US firms of the R&D budget that created the lead in the first place. You can debate whether AI chips are different because of military dual-use urgency, but dismissing Jensen as self-interested ignores a 40-year pattern.

16 4 replies
@UmaCienfuegos 2026-04-16

Mr. Patel- Superb interview! You pushed the questioning to the crux, to the premise that it is not, in fact, intelligent to send chips to China, because IN FACT it is not unless the AIM is something other than what it ought to be. Excerpts: 1:22:20, 1:24:31, 1:26:45. The AIM is clearly to keep the AI bubble growing, now encompassing the idea that working with China (sending chips) is highly beneficial. There is a huge benefit just be very CLEAR about what the benefit is and WHO PROFITS. Mr. Patel, what a superb interview. Extremely intelligent. You were not in the extremes, you were not the one raising straw men. Your questioning was very impeccable and well measured. Thank you.

8
@mark3xZod 2026-04-16

Obvious props go to Dwarkesh, but for this interview props to Jensen for talking thru genuinely tough questions. Instead of talking around the questions he was for most part willing to engage. Believing in yourself and your business enough to go deep on issues in your business is a rare trait. Elon doesn’t do it, Zuck walks around it and Nadella gets caved in when he tries to do it.

6
@desgreene2243 2026-04-16

This in retrospect will be quite an important interview.

5

Unlock the Data Inside
Turn Videos into Knowledge

  • Get FREE 10/day: transcripts, summaries, chats
  • Chat with videos, export text & PDF
  • $1 free API credit for RAG, chatbots & research

Free forever plan • All features unlocked

App screenshot