Guest
Dylan Patel
Dylan Patel , founder of SemiAnalysis , provides a deep dive into the 3 big bottlenecks to scaling AI compute: logic, memory, and power.
Appearance timeline
Follow this guest across the archive and see when each conversation happened.
2026
1 appearance
Dwarkesh Podcast / March 13, 2026 / Episode #104
Dylan Patel — Deep dive on the 3 big bottlenecks to scaling AI compute
Plus, why an H100 is worth more today than 3 years ago
Start here
Nvidia secured TSMC allocation early; Google is getting squeezed
Nvidia secured TSMC allocation early; Google is getting squeezed
2025
1 appearance
Lex Fridman Podcast / February 3, 2025 / Episode #459
DeepSeek, China, OpenAI, NVIDIA, xAI, TSMC, Stargate, and AI Megaclusters
Dylan Patel is the founder of SemiAnalysis, a research & analysis company specializing in semiconductors, GPUs, CPUs, and AI hardware. Nathan Lambert is a research scientist at the Allen Institute for AI (Ai2) and the author of a blog on AI called Interconnects. Thank you for listening ❤ Check out our sponsors: https://lexfridman.com/sponsors/ep459-sc See below for timestamps, transcript, and to give feedback, submit questions, contact Lex, etc.
Start here
DeepSeek-R1 and DeepSeek-V3
A lot of people are curious to understand China's DeepSeek AI models, so let's lay it out. Nathan, can you describe what DeepSeek-V3 and DeepSeek-R1 are, how they work, how they're trained? Let's look at the big picture and then we'll zoom in on the details. DeepSeek-V3 is a new mixture of experts, transformer language model from DeepSeek who is based in China. They have some new specifics in the model that we'll get into. Largely this is a open weight model and it's a instruction model like what you would use in ChatGPT. They also released what is called the base model, which is before these techniques of post-training. Most people use instruction models today, and those are what's served in all sorts of applications. This was released on, I believe, December 26th or that week. And then weeks later on January 20th, DeepSeek released DeepSeek-R1, which is a reasoning model, which really accelerated a lot of this discussion.
2024
1 appearance
Dwarkesh Podcast / October 2, 2024 / Episode #72
@Asianometry & Dylan Patel — How the semiconductor industry actually works
A bonanza on the semiconductor industry and hardware scaling to AGI by the end of the decade.
Start here
Architectures lead to different AI models? China vs. US
Architectures lead to different AI models? China vs. US
Where to start
The clearest entry points into this guest's appearances.
Profile
A short introduction based on public episode notes and linked profiles.
Dylan Patel
Dylan Patel , founder of SemiAnalysis , provides a deep dive into the 3 big bottlenecks to scaling AI compute: logic, memory, and power.
Appearances
3
Timeline span
October 2, 2024 to March 13, 2026
Social handles
Public profiles mentioned in the episode notes.
Podcasts
See which podcast archives this guest appears in most often.
Common topics
Themes that show up most often across this guest's appearances.