I note that NVIDIA structured a non‑exclusive license for Groq’s low‑latency inference technology, hired the engineering team and folded the chip into its serving platform as a specialized accelerator—Jensen framed it as ‘we’ll extend our architecture with Groq as an accelerator in very much the ways that we extended NVIDIA’s architecture with Mellanox.’ A quarterly‑filing footnote later disclosed about $3.96 billion of accrued purchase consideration, roughly $4 billion, the first hard quantification, comparable to both $2 billion laser stakes combined.

src ▸

NVDA-playbook / Absorbing the threat that beat you

rcpt ▸

/p/p-day-20260729-nvda-playbook-src-nvda-playbook-groq-rotation ↗

ts ▸

2026-07-29T07:33:15.471Z