Nvidia Unveils New AI Chip Powered by Groq Technology

Nvidia's $20 billion bet on Groq's chip tech aims to boost its AI computing ecosystem

Mar. 13, 2026 at 4:55pm

Got story updates? Submit your updates here. ›

Nvidia is expected to unveil a new AI chip at its upcoming GTC conference, leveraging technology licensed from startup Groq. The move is part of Nvidia's strategy to strengthen its position in the growing inference market for AI computing, where it faces increasing competition from rivals like AMD, Google, and Amazon. The new chip is designed to complement Nvidia's existing GPU-based offerings, with a focus on speed and efficiency for real-time AI workloads.

Why it matters

The AI computing market is rapidly evolving, with companies vying to provide the most powerful and cost-effective solutions for both training and inference. Nvidia's dominance in training has fueled its growth, but the inference market is more crowded. By incorporating Groq's specialized chip technology, Nvidia aims to bolster its inference capabilities and maintain its leadership in the AI computing space.

The details

Nvidia's $20 billion deal with Groq involves licensing the startup's technology and hiring key employees, including former CEO Jonathan Ross, who previously worked on Google's Tensor Processing Units (TPUs). Groq's chips, known as Language Processing Units (LPUs), are designed to excel at inference tasks by prioritizing speed and efficiency over the parallel processing capabilities that make Nvidia's GPUs well-suited for training. The integration of Groq's technology is expected to help Nvidia address the growing demand for cost-effective inference solutions as AI adoption continues to rise.

  • Nvidia's annual GTC conference is scheduled for next week (March 18-21, 2026) in San Jose, California.
  • Nvidia announced the $20 billion deal with Groq on Christmas Day 2025.

The players

Nvidia

An American technology company that designs graphics processing units (GPUs) for the gaming and professional markets, as well as system on a chip units (SoCs) for the mobile computing and automotive market.

Groq

A chip startup that developed specialized Language Processing Units (LPUs) focused on efficient inference processing for AI applications.

Jonathan Ross

The former CEO of Groq, who now holds the title of Chief Software Architect at Nvidia after joining the company as part of the Groq deal.

Google

The tech giant that previously employed Jonathan Ross, who was part of the team that developed Google's Tensor Processing Units (TPUs).

Advanced Micro Devices (AMD)

The second-largest maker of GPUs, which has found some traction in the inference market, recently signing up Meta Platforms as a customer.

Got photos? Submit your photos here. ›

What they’re saying

“I've got some great ideas that I'd like to share with you at GTC.”

— Jensen Huang, CEO, Nvidia (Nvidia Earnings Call)

“GPUs are really great at training models. When somebody wants to train a model, I'm just like, 'Just use GPUs. Don't talk to us.'”

— Jonathan Ross, Former CEO, Groq (Lumida Podcast)

“We're actually so crazy fast compared to GPUs that we've actually experimented a little bit with taking some portions of the model and running it on our LPUs and letting the rest run on GPU. And it actually speeds up and makes the GPU more economical.”

— Jonathan Ross, Former CEO, Groq (The Capital Markets Podcast)

What’s next

Nvidia is expected to provide more details on its plans for integrating Groq's technology into its AI computing ecosystem during the upcoming GTC conference.

The takeaway

Nvidia's $20 billion bet on Groq's specialized chip technology underscores the company's commitment to staying ahead in the rapidly evolving AI computing market, where it faces growing competition. The integration of Groq's expertise in efficient inference processing could help Nvidia strengthen its position and offer more comprehensive solutions to its customers.