Dylan PatelFounder and Chief Analyst, SemiAnalysis · The semiconductor-AI stack

Jensen giving out Pocky sticks in Asia

3 Nov 2025 · 0 min

Watch on YouTube

0:15

In this brief clip, Speaker A asks a presenter about any technological differences observed between two companies after their cooperation, and inquires about the current development stage and expected delivery timeline for HBM4 (High Bandwidth Memory 4). The question seeks clarification on post‑cooperation tech integration and the readiness of the next‑generation memory product.

More from Dylan

1:44

Sparse attention fixes memory bandwidth, not memory capacity.

8h ago

2:05

Inside the SemiAnalysis STEEL Lab

1d ago

2:05

STEEL: From product to transistor | SemiAnalysis Model Overview

1d ago

1:31

The AI buildout, quantified | SemiAnalysis

1d ago

0:43

Is GLM 5.3 worth sounding the alarms? 🚨

2d ago

44:13

Ep. 038 - Is GLM 5.3 a Mythos-Class Cyber Model From China? (InferenceX)

2d ago

2:04

Even the frontier labs underestimated their own demand.

5d ago

1:52

The AI buildout needs more capital than backstops can cover.

5d ago

1:52

The fastest finish is not always the most efficient one.

6d ago

1:23:18

Ep. 037 - Who's Funding the $11 Trillion AI Buildout? (Capital Markets)

1w ago

0:48

"It's like a zombie apocalypse, but you don't know how many waves of zombies are coming."

1w ago

See all content