← GPU HBM capacity and bandwidth

What podcasts say about GPU HBM capacity and bandwidth

Every statement, with the speaker, the exact quote and the moment it was said.

What experts have said about GPU HBM capacity and bandwidth

1 statement · 1 neutral

  1. Reiner PopeNeutralApr 29, 2026· Dwarkesh Podcast

    Memory capacity divided by bandwidth commonly yields an inference interval near 20 milliseconds.

    “that is capacity divided by bandwidth that tends to be 20 milliseconds”

    Listen at 24:10

    Open the episode · Reiner Pope – The math behind how LLMs are trained and served

Statements are attributed to the speaker as said on the episode and reflect their view at the time, not PodLume's. They are not advice.

GPU HBM capacity and bandwidth: what podcasts say · PodLume