On Wednesday (December 24) local time, Groq, an American chip design startup, announced on its website that it had partnered with Nvidia. A non-exclusive licensing agreement was reached, authorizing Nvidia to use Groq's inference technology.
Groq stated that the agreement reflects the shared commitment of both parties to expanding the application of high-performance, low-cost inference technologies.
The statement reads that, as part of the agreement, Groq founder Jonathan Ross, president Sunny Madra, and other members of the Groq team will join NVIDIA to help advance and scale the licensed technology.
The press release also stated that Groq will continue to operate as an independent company, with Simon Edwards serving as CEO. Apart from this, GroqCloud will continue its normal operations without interruption.
Shortly before Groq's official announcement, Alex Davis, CEO of the US technology investment firm Disruptive, revealed that Nvidia had agreed to acquire Groq for approximately $20 billion in cash.
In response, Nvidia stated that the company did not acquire Groq, but only made a technology licensing arrangement.
In the past two years, other large technology companies have also used this "quasi-acquisition" approach—not directly acquiring equity to avoid antitrust scrutiny—to compete for core technologies and talent in the AI field.
Disruptive has invested in Palantir and Airbnb . Well-known companies such as Spotify and Shield AI. In September of this year, Disruptive led Groq's latest funding round, raising $750 million and valuing the company at $6.9 billion post-money.
Davis stated that Disruptive has invested over $500 million in Groq since its founding in 2016. In his report, Davis revealed that Nvidia "acquired" Groq for approximately $20 billion in cash.
If true, this would be Nvidia's largest acquisition to date. Previously, Nvidia's largest acquisition was in 2019, when it acquired Israeli chip design company Mellanox for nearly $7 billion.
Nvidia also offered to acquire Arm for $40 billion, but the deal was blocked due to significant regulatory challenges.
As of the end of October, Nvidia held $60.6 billion in cash and short-term investments, up from $13.3 billion at the beginning of 2023. With its increasing cash reserves, Nvidia has been continuously increasing its investments in chip startups and the entire ecosystem in recent years.
In September of this year, Nvidia announced plans to invest up to $100 billion in OpenAI. That same month, Nvidia also announced plans to invest in Intel. Invest $5 billion.
In addition, Nvidia has invested in AI and energy infrastructure company Crusoe, AI model developer Cohere, and further increased its investment when AI cloud service provider CoreWeave went public.
Cailian Press introduced Groq back in February last year, when Groq launched a brand new AI chip. They claim to have achieved "the most powerful inference on Earth"—inference speeds on large models running on Groq are 10 times faster or even higher than those on NVIDIA GPUs.
Groq chips are understood to be a type of LPU (Language Processing Unit), primarily used to accelerate the inference-related tasks of large language models. Amidst the booming demand for AI acceleration chips, Groq's revenue target for this year is $500 million.
It's worth mentioning that Groq's current CEO, Jonathan Ross, was involved in developing Google's TPU (Tensor Processing Unit) before founding the company in 2016—a chip currently considered by many as one of the alternatives to Nvidia's GPUs.
Last month, the White House and the U.S. Department of Energy... The latest statement from the ministry shows that 24 top artificial intelligence companies The company has signed an agreement with the U.S. government to join Genesis . The plan includes Nvidia and Groq.
(Article source: CLS)
