Skip to content
gentic.news — AI News Intelligence Platform
Connecting to the Living Graph…

Listen to today's AI briefing

Daily podcast — 5 min, AI-narrated summary of top stories

A cluster of advanced GPU servers with cooling systems in a modern data center, illuminated by blue and green lights

Nvidia to Acquire Groq; Leftover Biz Buys Blackwell Clusters

Nvidia reportedly acquiring Groq; leftover Groq rents B300/GB300/Rubin GPUs, many Blackwell-only, per SemiAnalysis.

·3d ago·3 min read··23 views·AI-Generated·Report error
Share:
Is Nvidia acquiring Groq, and what is the leftover Groq doing with Blackwell GPUs?

Nvidia is reportedly acquiring Groq, per SemiAnalysis. The remaining Groq entity is buying Nvidia B300/GB300/Rubin GPUs and LPUs, renting them to enterprises and neolabs. Many clusters lack LPUs and are straight Blackwell, indicating a shift to GPU rental over LPU inference.

TL;DR

Nvidia reportedly acquiring Groq, per SemiAnalysis. · Remaining Groq renting B300/GB300/Rubin GPUs. · Many Groq clusters are Blackwell, not LPUs.

Nvidia is acquiring Groq, per @SemiAnalysis_. The leftover Groq will buy B300/GB300/Rubin GPUs and rent them out.

Key facts

  • Nvidia acquiring Groq per @SemiAnalysis_.
  • Leftover Groq buying B300/GB300/Rubin GPUs.
  • Many clusters are Blackwell, not LPU.
  • Deal would be Nvidia's biggest AI chip M&A since ARM.
  • No financial terms disclosed.
  • Nvidia data-center revenue $30.8B in Q3 FY2025.

Nvidia is reportedly acquiring Groq, the AI inference chip startup known for its LPU (Language Processing Unit), according to a post from @SemiAnalysis_. The deal, if confirmed, would give Nvidia control of Groq's LPU technology, a direct competitor to its own GPU-based inference stack.

The post describes a two-part structure: after the acquisition, the "leftover" Groq entity—presumably the non-core business—will purchase Nvidia's latest hardware, including B300, GB300, and Rubin GPUs, as well as LPUs, and rent these out to enterprises and "neolabs" (AI startups). According to @SemiAnalysis_, a significant portion of these clusters won't even include LPUs; they'll be straight Blackwell clusters.

This is a striking pivot. Groq built its brand on LPU inference speed, claiming 10x throughput over GPUs. If the remnant is now reselling Nvidia GPUs, it signals that the LPU advantage may be eroding in the market, or that Nvidia's acquisition is effectively neutralizing a competitor while monetizing its infrastructure.

No financial terms were disclosed, and neither Nvidia nor Groq has issued an official statement. The report is a single tweet, so treat it as unconfirmed. But the strategic logic is clear: Nvidia absorbs the technology, and the leftover entity becomes a GPU rental business, competing with the very cloud providers it once challenged.

The deal would mark Nvidia's most significant AI chip acquisition since its attempted purchase of ARM in 2020, which collapsed under regulatory pressure. Whether this one faces similar scrutiny remains to be seen, but the FTC has been active on AI M&A, as seen in its review of Microsoft's Inflection deal.

For enterprises, the implication is straightforward: if Groq's LPUs become Nvidia IP, the inference market consolidates further, leaving fewer non-Nvidia options. The rental business, meanwhile, feeds Nvidia's data-center revenue stream, which already hit $30.8 billion in Q3 FY2025.

Watch for official confirmation, details on the purchase price, and whether the FTC or EU regulators step in. Also track whether Groq's LPU roadmap survives the acquisition or gets shelved in favor of GPU expansion.

What to watch

The Groq Era Begins: Inside NVIDIA’s Largest-Ever Acquisition …

Watch for official confirmation from Nvidia or Groq, the purchase price, and regulatory filings with the FTC or EU. Also monitor whether Groq's LPU roadmap is sunset or integrated, and if the leftover rental business lands enterprise contracts that would show up in Q1 2026 earnings.

Source: gentic.news · · author= · citation.json

AI-assisted reporting. Generated by gentic.news from multiple verified sources, fact-checked against the Living Graph of 4,300+ entities. Edited by Ala SMITH.

Following this story?

Get a weekly digest with AI predictions, trends, and analysis — free.

AI Analysis

The report, while unconfirmed, fits a pattern of Nvidia acquiring or neutralizing inference competitors. Groq's LPU was a genuine threat in low-latency inference, but its market share remained niche. Buying it removes a technical rival and converts its infrastructure into a GPU rental business, which feeds Nvidia's core revenue. This mirrors Nvidia's earlier move with Cumulus Networks and Mellanox—acquire the technology, then fold it into the ecosystem. The 'leftover' renting Blackwell clusters is the telling detail. It suggests Groq's own hardware is not the primary revenue driver; instead, the residual entity becomes a cloud provider, competing with CoreWeave and others. This is a classic acqui-hire plus asset play, but the rental angle is unusual—it keeps the team intact and monetizes existing data-center contracts. Regulatory risk is real. The FTC has scrutinized AI acquisitions, and Nvidia's dominance in AI accelerators makes any deal a target. But given the size (Groq is private, likely valued under $10B), it may pass. The bigger question is whether Nvidia will continue to license LPU technology or bury it. Historical precedent suggests the latter, as seen with ARM's failed acquisition and Mellanox's integration.
This story is part of
The AI Infrastructure War Shifts from Chips to Developer Tools
Nvidia's enterprise pivot and AWS's OpenAI bet collide with Cursor's quiet ascent
Compare side-by-side
Nvidia vs Groq
Enjoyed this article?
Share:

AI Toolslive

Five one-click lenses on this article. Cached for 24h.

Pick a tool above to generate an instant lens on this article.

Related Articles

From the lab

The framework underneath this story

Every article on this site sits on top of one engine and one framework — both built by the lab.

More in Products & Launches

View all