Nvidia AI chip challenger Groq raises even more than expected, hits $6.9B valuation (techcrunch.com)

Groq confirmed a $750 million funding round at a $6.9 billion post-money valuation, topping earlier leaks and more than doubling its valuation since an August 2024 round that priced the company at $2.8 billion. The round was led by Disruptive with participation from BlackRock, Neuberger Berman, Deutsche Telekom Capital Partners and returning investors such as Samsung, Cisco and Altimeter. PitchBook estimates Groq has now raised over $3 billion to date. The company also reports rapid adoption — powering apps used by more than 2 million developers, up from 356,000 a year ago. The raise underlines growing investor confidence in Groq’s bid to challenge Nvidia’s GPU dominance with a distinct compute stack: LPUs (language processing units) and an “inference engine” designed to run AI models faster and cheaper than general-purpose GPUs. Groq sells both cloud services and on-prem racks made of integrated hardware/software nodes and supports open versions of major models (Meta, Qwen, Mistral, Google, OpenAI, etc.), making it attractive for enterprises seeking lower-cost, high-throughput inference or tighter data control. Founder Jonathan Ross’s TPU background lends credibility to the hardware-led approach. For the AI/ML community, Groq’s momentum signals increased competition in accelerator design, potential downward pressure on inference costs, and broader deployment choices for latency-, cost- or data-sensitive applications.