US-based AI chip startup Groq has announced it has secured $750 million through its latest funding round, placing the firm’s valuation at $6.9 billion.
Groq, an artificial intelligence (AI) chip startup, has secured $750 million in fresh funding, more than doubling its valuation in just under a year. After topping estimates of $600 million, the company is now valued at $6.9 billion, up from $2.8 billion last year.
The California-based firm has emerged as a new challenger to Nvidia, having positioned itself as a lower-cost alternative to dominant GPU-based systems in AI. Founded in 2016 by former Google engineer Jonathan Ross, Groq designs what it calls LPUs, or language processing units.
Company bosses claim these LPUs run inference workloads faster and with up to ten times higher energy efficiency compared with standard GPUs. Groq also uses a custom compiler called RealScale, which compresses models without significantly reducing output quality. This allows for smoother performance and helps reduce costs for businesses and developers running AI models.
Who led Groq’s latest round of funding?
The round of funding was led by Disruptive, a Dallas-based growth investment firm that has backed major tech companies such as Palantir and Spotify. Groq confirmed on Wednesday that Disruptive invested around $350 million.
Meanwhile, other significant investments came from BlackRock, Neuberger Berman, Deutsche Telekom, Samsung, and Cisco. The round also included Samsung, D1, Altimeter, 1789 Capital, and Infinitum. Groq has now raised more than $3 billion in total funding, after securing $640 million in August 2024 at a $2.8 billion valuation.
This freshly raised capital will enable Groq to expand its global infrastructure, including the construction of new data centres in different regions, including its first Asia-Pacific facility in 2025. The company currently operates 13 facilities across North America, Europe, and the Middle East, supporting clients such as Saudi Arabia’s Humain.
Groq set to become a serious AI player
Ongoing demand for Groq’s computing power continues to outpace supply, with Ross confirming customers are requesting more capacity than the company can currently deliver. Groq recently increased capacity by more than 10% in just one month, but all of that was quickly absorbed.
Through its existing business model, the AI firm sells products as GroqRack hardware or through GroqCloud, giving developers the choice of on-premises installations or cloud-based services. Both options support open models from Meta, Google, OpenAI, Mistral, DeepSeek, and other major technology players in further advancing the AI industry.
Earlier this year, Saudi Arabia pledged $1.5 billion to secure Groq’s advanced chips as part of plans to boost domestic AI infrastructure. The contracts are expected to help generate about $500 million in revenue this year, strengthening Groq’s global market position.
Groq believes inference, where trained AI models run in real-world applications, will define the next era of AI growth. While Nvidia remains the market leader of training chips, Groq aims to challenge its position in inference with high-speed, predictable, and lower-cost alternatives.
Remember to stay updated with all the latest Market News at FXTrustScore.com.
