SemiAnalysis breaks it down, one chip at a time

2 Oct 2025 · 0 min
0:29

In this brief analyst commentary, the speaker discusses Nvidia's cloud business focus, highlighting the recent acquisition of Lepton. Lepton is described as a platform‑as‑a‑service (PaaS) that operates at the lower layer of Nvidia's NeoCloud ecosystem, allowing commodity‑infrastructure providers to deliver DGX‑level GPU services. By integrating Lepton as the underlying platform, Nvidia is moving away from its proprietary DGX Cloud offering and instead enabling a bottom‑up approach where NeoClouds can sell GPU capacity directly to customers. The segment emphasizes Lepton's role as an enabler for broader GPU market penetration through software‑driven infrastructure.

More from Dylan

1:44
youtube.com

Sparse attention fixes memory bandwidth, not memory capacity.

10h ago
2:05
youtube.com

Inside the SemiAnalysis STEEL Lab

1d ago
2:05
youtube.com

STEEL: From product to transistor | SemiAnalysis Model Overview

1d ago
1:31
youtube.com

The AI buildout, quantified | SemiAnalysis

1d ago
0:43
youtube.com

Is GLM 5.3 worth sounding the alarms? 🚨

2d ago
44:13
youtube.com

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

2d ago
2:04
youtube.com

Even the frontier labs underestimated their own demand.

5d ago
1:52
youtube.com

The AI buildout needs more capital than backstops can cover.

6d ago
1:52
youtube.com

The fastest finish is not always the most efficient one.

6d ago
1:23:18
youtube.com

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

1w ago
0:48
youtube.com

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

1w ago
See all content