Groq, a startup focused on developing chips to accelerate generative AI models, announced on Monday that it has secured $640 million in a new funding round led by Blackrock. Other participants included Neuberger Berman, Type One Ventures, Cisco, KDDI, and Samsung Catalyst Fund.

This funding round, which brings Groq’s total raised to over $1 billion and values the company at $2.8 billion, is a significant achievement for the startup. Initially, Groq aimed to raise $300 million at a $2.5 billion valuation. This new funding more than doubles Groq’s previous valuation of approximately $1 billion in April 2021, when it raised $300 million in a round led by Tiger Global Management and D1 Capital Partners.

Meta’s chief AI scientist, Yann LeCun, will join Groq as a technical advisor, and Stuart Pann, former head of Intel’s foundry business and ex-CIO at HP, will become the chief operating officer. LeCun’s involvement is notable given Meta’s own investments in AI chips, providing Groq with a strong ally in a competitive market.

Founded in 2016, Groq is developing an LPU (language processing unit) inference engine, which it claims can run generative AI models similar to OpenAI’s ChatGPT and GPT-4 at ten times the speed and one-tenth the energy consumption.

Groq’s CEO, Jonathan Ross, is known for co-inventing Google’s tensor processing unit (TPU), a custom AI accelerator chip. Ross co-founded Groq with Douglas Wightman, a former engineer at Alphabet’s X moonshot lab.

Groq offers an LPU-powered developer platform called GroqCloud, which includes models like Meta’s Llama 3.1, Google’s Gemma, OpenAI’s Whisper, and Mistral’s Mixtral. The platform also provides an API for customers to use its chips in cloud instances. GroqCloud, which launched late last year, now has over 356,000 developers. A portion of the new funding will be used to scale capacity and add new models and features.

Groq faces increasing competition from both AI chip startups and established players like Nvidia, which controls a significant share of the AI chip market. Nvidia is taking aggressive steps to maintain its dominance, including releasing new AI chip architectures annually and establishing a business unit for bespoke chip designs.

In addition to Nvidia, Groq competes with Amazon, Google, and Microsoft, all of which offer custom chips for AI workloads. Groq also faces competition from Arm, Intel, AMD, and other startups in an AI chip market that could reach $400 billion in annual sales within the next five years.

To carve out its niche, Groq is investing in enterprise and government outreach. In March, Groq acquired Definitive Intelligence to form a new business unit called Groq Systems, which serves organizations including U.S. government agencies. Groq has also partnered with Carahsoft to sell its solutions to public sector clients and has a letter of intent to install its LPUs at Earth Wind & Power’s Norway data center. Additionally, Groq is collaborating with Aramco Digital to install LPUs in future data centers in the Middle East.

Groq is also advancing its chip technology. The company announced last August that it would contract with Samsung’s foundry business to manufacture 4nm LPUs, expected to deliver performance and efficiency gains over its first-generation 13nm chips. Groq plans to deploy more than 108,000 LPUs by the end of Q1 2025.