Episode

Designing the AI Factory from the Rack Up

Podcast
EE Times Current
Published
Aug 21, 2026
Duration seconds
1102
Processing state
not_requested
Canonical source
https://podcast.show/aspencore/episode/155058884/
Audio
https://media.blubrry.com/aspencore/content.blubrry.com/aspencore/Aspencore_Synopsys_EET_Neeraj_Rev_1_081726.mp3
JSON
/v1/public/podcasts/ee-times-current-709434/episodes/designing-the-ai-factory-from-the-rack-up
Markdown
/podcast/ee-times-current-709434/designing-the-ai-factory-from-the-rack-up.md

Actions

  • POST https://stenobird.com/v1/public/podcasts/ee-times-current-709434/episodes/designing-the-ai-factory-from-the-rack-up/transcription-requests
    Idempotently request low-priority transcript generation for this episode.
  • GET https://stenobird.com/podcast/ee-times-current-709434/designing-the-ai-factory-from-the-rack-up.md
    Read the agent-friendly Markdown representation of this episode resource.

Summary

For years, AI infrastructure was measured at the scale of the data center. That's changed. In this episode, Neeraj Paliwal explains why the rack has become the indivisible unit of a modern AI system — designed as one system rather than a collection of servers — and why performance is no longer defined by the compute in any single box, but by how efficiently data moves, is stored, and is secured across the entire rack. And getting there depends on co-designing the silicon, memory, interconnect, and packaging together, with the IP foundation turning concentrated power into usable compute.