AIAIAI.FYI
LIVE

Who's who · Chips and infrastructure

Groq

Groq designed the LPU, a chip for fast AI inference, and now runs GroqCloud, an inference cloud service.

Why it matters: Nvidia paid a reported $20 billion to license its technology and hire its leaders, a sign of how valuable fast inference has become.

Key facts

  • Signed a non-exclusive licensing deal with Nvidia in December 2025, reportedly worth about $20 billion; founder-CEO Jonathan Ross and president Sunny Madra joined Nvidia, and Groq said it would stay independent.Source: en.wikipedia.org, as of Dec 2025
  • Was reported in May 2026 to be raising $650 million from existing investors to fund its shift to an inference cloud business.Source: en.wikipedia.org, as of May 2026
  • Describes itself as a neocloud for fast inference building hundreds of megawatts of capacity, with an 'LPX' product that works alongside Nvidia GPUs.Source: groq.com, as of Nov 2026

Our latest stories

No stories about Groq in our recent coverage yet. New ones appear here as they publish.