NVIDIA, every mention
26 scenes (2026), the whole family · ← back to NVIDIA
tap a year for its mentions
every year 2026 anyone Garry Tan 19Jensen Huang 15Stuart (Stu) 6Jared Friedman 6Elon Musk 5Ankit Gupta 3Francois Chaubard 2Diana Hu 2Pedro Franceschi 1Bob McGrew 1
Verbatim, from the transcripts: the passages where NVIDIA comes up
Multi-GPU Kernels, Intelligence per Watt, Heterogeneous Inference, and More | YC Paper Club · Y Combinator
- ▶ 2:33 Francois Chaubard And then Misha, uh, hardware software co-design, been working on that for two decades and ran AI infra at NVIDIA, uh, and then also worked on hardware software co-design at Meta.
- ▶ 4:05 Francois Chaubard People will go to, um, uh, Nvidia for pre-fill and then they'll go to cerebris for, for decode engine.
- ▶ 12:02 Stuart (Stu) Um, on Blackwell GPUs, you have one 48 SMs.
- ▶ 12:05 Stuart (Stu) On Hopper, you have one 32 SMs.
- ▶ 12:45 Stuart (Stu) It's relatively slow, so we have NVLink, which allows direct communication of data between the GPUs, and that's basically essentially, um, what NVIDIA DGX machines are all about. 2 times in the scene
- ▶ 15:27 Stuart (Stu) Um, so on Blackwell, we find that with roughly 15 SMs out of one 48, um, TMA is able to saturate the NVLink.
- ▶ 20:57 Stuart (Stu) For example, Cursor is using it to train Composer on tens of thousands of Blackwell GPUs.
- ▶ 22:11 unnamed speaker Um, and that's kind of how we're still constructing at least the data centers today, where you have these warehouses filled with, um, TPUs or, um, GPUs from, uh, Google and Nvidia respectively. 4 times in the scene
- ▶ 25:59 unnamed speaker Over the past two to three years, and so this is from Apple, from Nvidia, from AMD, from Zambanova, to basically understand what was the kind of upper bounds for intelligence per watt and intelligence per joule, um, and we wanted to do… 2 times in the scene
- ▶ 27:25 unnamed speaker Um, and the actual intelligence delivered per joule, and this is primarily driven by better accelerators, and so this is by more memory being placed on, um, consumer GPUs, uh, such as the Apple, um, M four Macs, um, as well as the, um, uh,…
- ▶ 29:12 unnamed speaker Um, just want to thank, um, our collaborators on this project, um, from NVIDIA, from Google, from Apple, AMD, Open Router, SambaNova.
- ▶ 33:16 unnamed speaker Um, but recently, especially starting with Blackwell, people are like, well, you know, the Triton's programming model is like very restrictive.
- ▶ 33:30 unnamed speaker Cutlass and QDSL, which are closed source libraries by NVIDIA for, like, really state-of-the-art matmoles.
- ▶ 38:05 unnamed speaker There's like a wonderful algorithm, you know, like NVIDIA would love to hire you.
Jensen Huang: The Mindset That Built NVIDIA · Y Combinator
- ▶ 0:15 Garry Tan Please join me in welcoming you to the stage, the founder and CEO of NVIDIA, Jensen Huang.
- ▶ 1:08 Garry Tan Uh, well, for the students who only know NVIDIA as a company at the center of AI, uh, what part of the early NVIDIA story do they most need to understand? 10 times in the scene
- ▶ 4:40 Jensen Huang And so, since then, NVIDIA has been, you know, inventing all kinds of technology since, all kinds of technology we've never, never really done before, and we approach everything with the same attitude, you know, this is, ah, if it's… 4 times in the scene
- ▶ 8:06 Jensen Huang But because our algorithm and our technology was fundamentally flawed, I went to Japan, and I told Irimandrisan, the CEO at the time, that, that the contract that they gave us was, like, a twelve million dollars contract. 8 times in the scene
- ▶ 14:08 Garry Tan Cause if the fortune 500 did that, like the fortune 500 will probably look a lot more like Nvidia than not.
- ▶ 23:48 Garry Tan Is that part of the thrust behind NVIDIA being so involved? 4 times in the scene
- ▶ 27:55 Garry Tan I feel like there's this pattern that I'm starting to see around NVIDIA. 2 times in the scene
- ▶ 34:05 Garry Tan Another really exciting thing that NVIDIA is all the way out on the edge on is actually physical robots.
- ▶ 37:54 Jensen Huang And so, uh, inside Waymo, uh, our chips from NVIDIA, uh, at, at Tesla, we were in the car, uh, now we're in the data center.
- ▶ 44:33 Jensen Huang I could tell you exactly how I felt when I first, when NVIDIA first founded and, and, uh, the three of us started, um, 2 times in the scene
World Models, JEPA And The Path To Sample-Efficient RL · Y Combinator
- ▶ 53:56 unnamed speaker Um, I think a bunch of companies, um, NVIDIA, uh, uh, this, this paper here, uh, is basically talking about doing exactly the same, this dream zero.
The Most AI-Pilled CEO We Know · Y Combinator
- ▶ 7:27 Pedro Franceschi So a lot of folks were, you know, and we saw NVIDIA and others on Nemo Claw, let's build these like open shell forks that have controls over, you know, what tools the model calls.