Q&A with Jensen Huang on Nvidia's supply chain moat, competition from ASICs like Google's TPU, investing in AI labs and neoclouds, selling to China, and more
“If our next several years are a trillion dollars in scale, we have the supply chain to do it”
Dwarkesh PodcastDwarkesh Patel
Context & Ripple Effects
Earlier coverage traced Nvidia’s position from integrated hardware and software, through DGX Cloud and the “AI factory” model, to its current emphasis on scaling AI infrastructure. This interview extends that arc from product platform to the supply-chain capacity behind it.
The discussion also places Nvidia’s expansion alongside custom silicon competition from Google’s TPU, direct investment in AI labs and neoclouds, and the continuing strategic importance of China as a market.
First-order effects
Nvidia is signaling that supply-chain access is central to its ability to support much larger AI-system deployments, not merely to ship successive chips.
Its investments in AI labs and neoclouds deepen Nvidia’s ties to customers and operators, while TPU competition makes custom accelerators a more explicit part of the market landscape.
Second-order effects
AI labs and neoclouds backed by Nvidia may gain earlier or more dependable access to Nvidia-based capacity, increasing pressure on other cloud and infrastructure providers to secure comparable supply or differentiate with alternative hardware.
Google and other ASIC backers have added incentive to prove that specialized accelerators can lower cost or improve workload performance enough to offset Nvidia’s broader platform and supply advantages.
Third-order effects
If supply availability remains a binding constraint, advantage in AI may increasingly accrue to firms that coordinate chips, networking, software, capital, and deployment channels rather than to chip designers alone.
The combination of custom-silicon rivalry, Nvidia’s ecosystem investments, and China exposure points toward a more strategically segmented compute market, where access and platform alignment matter alongside raw accelerator performance.
The trend: AI compute is becoming a vertically coordinated infrastructure business in which supply-chain control, ecosystem financing, and specialized silicon jointly shape competitive power.
It is as if you didn't listen to the interview. This is about ensuring the US tech stack is the gold standard around the world. Blatant misinterpretation is choice my man. [video]
Fascinating critique coming from Crowdstrike, which infamously helped to architect the Russia collusion hoax. Btw, Jensen specifically says here that his goal is for America to win.
Incredible interview with Jensen. He blatantly admits that his jihad against export controls is simply all about Nvidia selling more chips worldwide, not about national security or winning the AI race against China (which he previously said doesn't even matter if we win)
Jensen Huang isn't worried about big AI companies using non-Nvidia chips: “If they don't try these other things, how would they know how good ours is? Sometimes you gotta be reminded of it.” That's a different kind of confidence. cc: @nvidia, @dwarkesh_sp, @AnthropicAI, [video]
Dwarkesh not buying the Nvidia's talking points they are pushing through DC about selling chips to China...he seems frustrated at the skepticism...China bit starts at about 57:30
You rarely see a host keep pressing a guest this hard until the truth starts to come out (this is usually the best part of @dwarkesh_sp podcast...) But honestly, even after 30 minutes of back-and-forth, it still felt like they weren't really on the same page on China issue😂
There is a key theme here from Jensen that underscores my recurring thesis I continually unpack @DiligenceStack: The market evolves from a semiconductor scarcity story to an industrial systems story in which energy, networking, packaging, software, and algorithmic efficiency all
I asked Jensen: “2 out of the top 3 models in the world, Claude and Gemini, were trained on TPU. What does that mean for Nvidia going forward?” After a long technical back and forth about what the right accelerator for AI looks like (see full episode), Jensen lays down the [video…
@dwarkesh_sp It's so refreshing to watch someone who knows how to make an argument, most people working in media forgot how to argue. You really pressed him on China.
The Jensen Huang episode. 0:00:00 - Is Nvidia's biggest moat its grip on scarce supply chains? 0:16:25 - Will TPUs break Nvidia's hold on AI compute? 0:41:06 - Why doesn't Nvidia become a hyperscaler? 0:57:36 - Should we be selling AI chips to China? 1:35:06 - Why doesn't Nvidia …
Dwarkesh asks what's actually bottlenecking AI capacity right now. Huang's first answer isn't fabs. Isn't memory. Isn't power. Isn't even regulation. It's plumbers and electricians. I think Huang is more in line with how economists think about this than how AI people do. In [imag…
Nvidia CEO Jensen Huang's interview with @dwarkesh_sp is extraordinary and embarrassing. When he actually faces tough questions, Huang becomes a dishonest angry bully. Just a pathetic illustration of how bad our corporate leadership class is. https://www.youtube.com/...
Jensen apparently was also unintentionally making the case *for* export controls on Dwarkesh: “They [China] have datacenters that are sitting completely empty, fully powered. You know they have ghost cities, they have ghost datacenters too.” Imagine if China had the chips!
Jensen on China: “The amount of energy they have is incredible. Isn't that right? AI is a parallel computing problem, isn't it? Why can't they just put 4x, 10x, as many chips together because energy's free? They have so much energy. They have datacenters that are sitting
Jensen falling for the trope: “AI will commoditize a lot of things. Your things. My thing is extremely hard and it's out of reach for AI.” Great thing all the labs have ASICs projects, so we'll discover who's right.
@firstadopter I play devil's advocate against my guests to test out their ideas. As you pointed out yourself when I hassled Dario during *his* interview :) Re China, when I had Dario on, because he supports export controls, I asked him, “Why can't China and America both have a da…
The @dwarkesh_sp and Jensen podcast is a great listen. it illustrates the schism in a lot of industry debates around “AGI” and the concept of AI timelines. Both parties represent a different school of thought and this is a great distillation of a lot of offline debates.
It made for good entertaining pod and maybe he's just playing devil's advocate but @dwarkesh_sp was way too Anthropic point-of-view pilled (AI is uranium/nuclear weapons! Mythos! Zero day!)