Nvidia Competitor, Groq, Raises $750 Million in Funding 

Updated:September 19, 2025

Reading Time: 2 minutes
An AI chip laying beside a tall stack of cash

AI chip startup Groq has raised $750 million in new funding, placing the company at a post-money valuation of $6.9 billion. 

This figure exceeds earlier reports that estimated a raise of $600 million at a $6 billion valuation.

Rapid Growth

In August 2024, the company secured $640 million at a valuation of $2.8 billion. In just over a year, its valuation has more than doubled.

According to PitchBook, Groq has now raised more than $3 billion in total. Such momentum places the company among the most closely watched challengers to Nvidia in the AI chip sector.

AI Chips


A Groq AI chip
Source: Groq

Nvidia dominates the industry with its GPUs (graphics processing units). These chips are the standard for training and running AI models.

However, Groq is pursuing a different path. Instead of GPUs, it has developed LPUs (language processing units). These are designed for inference- the stage where AI models deliver results.

Groq calls its product an inference engine. It focuses on running AI workloads with high efficiency and reduced cost.

The company offers its technology through two main options: cloud services for immediate access, and on-premises clusters, which consist of server racks equipped with Groq’s integrated hardware and software.

Both formats support popular open-source AI models, including those from Meta, OpenAI, Google, Mistral, DeepSeek, and Qwen. 

Groq asserts that its systems maintain or even improve performance compared to alternatives while lowering costs.

Investors

The new round was led by Disruptive, a firm known for backing ambitious technology ventures.

Other participants included BlackRock, Neuberger Berman, and Deutsche Telekom Capital Partners.

Several existing investors also took part, such as Samsung, Cisco, D1, and Altimeter. 

Proven Experience

Groq’s founder, Jonathan Ross, brings significant expertise to the company. At Google, he helped develop the Tensor Processing Unit (TPU). 

Google introduced the TPU in 2016, the same year Groq emerged from stealth. TPUs remain a key part of Google Cloud’s AI services today.

Ross’s background gives Groq credibility in a field where technical expertise is critical.

Developer Adoption

Groq reports that more than 2 million developers now use its technology. This is a significant increase from 356,000 developers a year earlier.

Such growth highlights demand for alternatives to Nvidia’s chips, which are often costly and difficult to secure.

Lolade

Contributor & AI Expert