Tuesday, June 23, 2026
EN · DarkArchiveSubscribe

Slicast

AI Infrastructure · News & Analysis
HomeCompute & CloudReport
Compute & Cloud · Report

AI inference startup Groq raises $650 million Series D funding, re-staffing after Nvidia licensed IP and recruited key executives.

Validates independent inference accelerator market survives despite Nvidia talent acquisition; demonstrates investor confidence in non-Nvidia alternatives to training/inference split.
Trade pressSlicast · June 23, 2026 06:48 · US · Source: Google News
importance 92
Hero image 16:9 · placeholder
Image / Slicast · Source: GNews/global: AI inference (cost OR chip OR Groq OR Cerebras OR SambaNova)

Groq confirmed a $650 million funding round, which followed Nvidia striking a non-exclusive licensing agreement for Groq's technology and recruiting key executives including founder and CEO Jonathan Ross and President Sunny Madra, along with other employees. The funding round occurred roughly six months after Nvidia secured the licensing deal and recruited these executives.

Groq did not disclose a new valuation; its previous valuation stood at $6.9 billion following a $750 million round in September.

Ross, who previously worked at Google and helped create the Tensor Processing Unit (TPU), co-founded Groq more than ten years ago with Google engineer Doug Wightman. Wightman remained with the company after the Nvidia deal and became CEO.

Groq developed the Language Processing Unit (LPU), a chip designed for inference, which the company sold as part of a cloud service or as hardware for on-premises clusters. Following Nvidia's licensing of the LPU IP, the GPU maker announced its own inference infrastructure cluster—Nvidia Groq 3 LPX—at its March GTC event.

In response, Groq shifted its focus to its neocloud business, a direction previously led by Madra following Groq's 2024 acquisition of Definitive Intelligence. The neocloud operation now spans 13 data centers across North America, Europe, the Middle East, and the Asia-Pacific region, serving more than five million developers and thousands of AI companies while processing trillions of tokens weekly, according to Groq.

Groq also refreshed its leadership team. Alan Rice was named COO, having previously worked at xAI and Meta after serving in the U.S. Navy. Sinclair Schuller joined as CTO and Rakesh Malhotra as CPO. All three had previously worked together at Apprenda and later co-founded Nuvalence, which EY acquired in 2024. Malhotra spent more than ten years on Microsoft's cloud products.

Groq's success depends on whether its inference cloud service can remain competitive now that its core hardware IP is shared with Nvidia. While demand for inference technology continues to grow, the company faces intensifying competition and emerging innovations.

Other market players demonstrate that such transitions can succeed. Jason Droge, CEO of Scale AI, told Forbes that his company recovered after Meta's talent acquisition of roughly $14.3 billion about a year ago, and Scale AI expects approximately $1 billion in revenue.

The dynamics surrounding Groq illustrate broader market trends: companies can adapt to strategic changes, develop competitive cloud services, and maintain competitiveness even with shared IP alongside major competitors. The industry continues to offer room for various competitive scenarios.

Read the original ↗
AI inference startup Groq raises $650 million Series D funding, re-staffing after Nvidia licensed IP and recruited key executives. · Slicast