The Green Gambit

Nvidia committed $26 billion over five years to building open-weight AI models. This episode examines the strategy behind that bet: open weights as hardware lock-in, the Nemotron Coalition, NemoClaw agent runtime, the Vera Rubin and Feynman hardware roadmaps, and what it means that a chip company is now competing directly with AI labs on model quality.

Sources

Primary Reporting

• Nvidia Will Spend $26 Billion to Build Open-Weight AI Models, Filings Show — Wired, Will Knight, March 11 2026
• NVIDIA GTC 2026: Live Updates on What's Next in AI — Nvidia Blog, March 16–18 2026
• Leading Inference Providers Cut AI Costs by up to 10x With Open Source Models on NVIDIA Blackwell — Nvidia Blog

Models & Research

• Nemotron 3 Super — Research Page — Nvidia Research
• NVIDIA Expands Open Model Families — Nvidia Newsroom
• NVIDIA Launches Nemotron Coalition — Nvidia Newsroom

Hardware

• NVIDIA Vera Rubin Platform — Nvidia Newsroom
• NVIDIA Vera CPU — Nvidia Newsroom
• DGX Station GB300 — Nvidia

Agents & Stack

• NVIDIA NemoClaw — Nvidia Newsroom
• NVIDIA Agent Toolkit / OpenShell — Nvidia Newsroom
• GTC Spotlights NVIDIA RTX PCs and DGX Sparks Running Latest Open Models — Nvidia Blog

Context

• ATOM Project — American Truly Open Models — Allen Institute for AI
• Laude Institute — Nonprofit for AI openness
• DeepSeek R1 — January 2025 — Wired
• OpenAI GPT-OSS — Wired