companies

Groq Inc.

Designs the LPU (Language Processing Unit), a custom AI inference chip, and operates GroqCloud, a low-latency AI inference cloud platform.

Groq, Inc. was founded in 2016 by former Google engineers led by Jonathan Ross, one of the original designers of Google's Tensor Processing Unit (TPU), along with Douglas Wightman (ex-Google X). Headquartered in Mountain View, California, Groq builds an AI accelerator chip, originally the Tensor Streaming Processor and later rebranded the Language Processing Unit (LPU), and operates GroqCloud, a developer platform launched in February 2024 for low-latency LLM inference. Groq raised a $10M seed in 2017 (Social Capital), a $300M round in 2021 (Tiger Global, D1 Capital), a $640M Series D in August 2024 (BlackRock; $2.8B valuation), and a $750M round in September 2025 at a $6.9B valuation. In December 2025, Nvidia and Groq announced a non-exclusive technology licensing agreement reported at roughly $20B, under which founder Jonathan Ross and other senior staff moved to Nvidia while Groq remained independent; Douglas Wightman became CEO. In June 2026, Groq confirmed a further $650M raise to fund its pivot toward an AI inference cloud business, reporting 13 data centers serving over 5 million developers.

Founded

2016

Headquarters

Mountain View, United States

Sector

infrastructure