Key Takeaways
- Federal investigators are examining whether Nvidia’s $20B Groq licensing agreement was structured to circumvent antitrust oversight.
- Under the December arrangement, Nvidia obtained licensing rights to Groq’s technology while the startup retained its independence, though its CEO and COO transitioned to Nvidia.
- Congressional members have condemned the transaction structure, arguing it functions as a de facto acquisition that undermines competition in AI semiconductor markets.
- The Groq 3 LPX rack has entered full-scale production, achieving 3,400 tokens per second performance, with deployment scheduled at neocloud Nebius this year.
- Federal investigators may conclude their inquiry without pursuing enforcement measures, while Nvidia maintains the transaction advances technological progress.
Nvidia (NVDA) stock declined 0.91% following revelations that federal investigators are scrutinizing the semiconductor giant’s $20 billion licensing deal with artificial intelligence chip developer Groq.
Federal authorities are examining whether the transaction’s structure was deliberately designed to avoid mandatory antitrust evaluation. The New York Times initially disclosed the investigation, with Bloomberg subsequently verifying the confidential inquiry.
The transaction was revealed in December. According to its framework, Nvidia secured licensing privileges to Groq’s intellectual property while Groq maintained its operational autonomy. Groq’s chief executive Jonathan Ross and chief operating officer Sunny Madra both transitioned to Nvidia positions as part of the agreement.
Groq declined to file the transaction for mandatory antitrust examination, attracting regulatory interest. Federal questioning regarding the arrangement commenced earlier this year within a comprehensive antitrust investigation targeting Nvidia.
Congressional representatives have criticized the transaction framework, characterizing it as a functional acquisition designed to diminish competitive dynamics in artificial intelligence semiconductor markets. Nvidia has vigorously defended the arrangement.
“The Groq story is a prime example of the American system working as designed to promote innovation, reward entrepreneurs, and benefit consumers,” an Nvidia spokesperson said.
The federal inquiry may conclude without resulting in enforcement proceedings.
Groq Production System Deployed
Regarding product developments, the Groq 3 LPX rack has achieved full production status and is scheduled for installation at neocloud Nebius before year-end. Each liquid-cooled system incorporates 256 Groq processors and integrates with Nvidia’s Vera CPUs and Rubin GPUs.
Nvidia reports the system achieves 3,400 tokens per second throughput, significantly exceeding the 750 tokens per second OpenAI has announced for its Cerebras-powered Ultrafast configuration. Groq’s processors are fabricated by Samsung, contrasting with Nvidia’s GPUs manufactured by Taiwan Semiconductor Manufacturing.
The $20 billion transaction represents Nvidia’s largest acquisition to date. Bernstein analyst Stacy Rasgon informed CNBC that Nvidia possesses sufficient financial resources to execute a transaction of this magnitude without compromising its financial position.
Institutional Investors Maintain Confidence
Nvidia appeared in 285 hedge fund portfolios during Q2 2026, increasing from 275 in Q1. Competitor AMD was present in 164 hedge fund holdings during the comparable timeframe.
Nebius remains the sole confirmed cloud infrastructure provider scheduled to implement the Groq rack system. Nvidia is simultaneously pursuing multiple significant strategic investments, including financial commitments reaching $100 billion to OpenAI and $5 billion to Intel.
The Department of Justice investigation into the Groq licensing transaction remains active.



