TBPN

NVIDIA Earnings Breakdown | Doug O’Laughlin, Ajay Agarwal, Koen Bok & Jorn Van Dijk, Hussein Fazal, Oisin Hanrahan, Jon Callaghan, Shane Hegde

62 snips
Aug 28, 2025
In this engaging discussion, Doug O’Laughlin, President of SemiAnalysis, shares insights on NVIDIA's recent earnings and the complexities of AI investments. Ajay Agarwal, of Bain Capital Ventures, reflects on software innovation and network effects. Shane Hegde, CEO of Air, emphasizes streamlined creative operations while Oisin Hanrahan of Keychain highlights AI's impact on the CPG supply chain. They tackle market challenges and explore the future of tech, with humor and unique anecdotes enriching the conversation.
Ask episode
AI Snips
Chapters
Books
Transcript
Episode notes
ADVICE

Watch The Data Center Line Closely

  • If you're evaluating NVIDIA's revenue, focus on data center growth and backlog specifics (e.g., H20/H200 license timing).
  • Read earnings detail, not headlines: small mismatches in the data center line move markets more than peripheral segments.
INSIGHT

CUDA Is Hard To Replace Because It's A System

  • Re-platforming ML stacks off CUDA is technically feasible but nontrivial due to deep software, tooling, and systems integration.
  • Unseating NVIDIA will require solving hard system-level bugs, not just re-compiling models.
INSIGHT

Cheaper Tokens But Exploding Token Usage

  • The economics of LLMs show inference cost per token falling but token consumption exploding, so total compute spend can still grow rapidly.
  • Product design must optimize for longer, token-heavy user workflows, not just cheaper per-token costs.
Get the Snipd Podcast app to discover more snips from this episode
Get the app