Updates
1. Serve Qwen3.8-2.4T-A95B, a 2.4T-Parameter Model, with Configurable Reasoning on NVIDIA GB300 NVL72
Serve Qwen3.8-2.4T-A95B, a 2.4T-Parameter Model, with Configurable Reasoning on NVIDIA GB300 NVL72.

Proof: NVIDIA AI Blog title names 2.4T-parameter model scale on the NVIDIA GB300 NVL72.
Key number: 2.4T
Impact: NVIDIA sets the tone for whether hyperscaler demand is still broadening or just pausing.
Watch next: Watch orders, backlog, and hyperscaler capex for signs that the buildout is still expanding across Semiconductors, AI Infrastructure.
Canonical host: developer.nvidia.com