TechCrunch News 09月18日
AI芯片初创公司Groq融资7.5亿美元,估值达69亿美元
index_new5.html
../../../zaker_core/zaker_tpl_static/wap/tpl_guoji1.html

 

AI芯片初创公司Groq近日宣布成功融资7.5亿美元,公司估值达到69亿美元。此次融资额远超此前传闻,标志着Groq在一年内估值翻倍。Groq致力于打破英伟达在AI芯片领域的垄断地位,其开发的语言处理单元(LPU)是一种优化的推理引擎,专注于快速高效地运行AI模型。Groq提供的云服务和本地部署硬件方案,能够以更低的成本提供与竞品相当甚至更优的AI性能。公司创始人Jonathan Ross曾参与谷歌TPU的开发,为Groq的技术积累奠定了基础。目前,Groq已服务超过200万开发者。

🚀 **巨额融资与估值跃升**:AI芯片初创公司Groq已确认完成新一轮7.5亿美元融资,公司投后估值高达69亿美元。此轮融资不仅超出了此前市场传闻的6亿美元,更重要的是,相较于2024年8月28亿美元的估值,Groq在不到一年的时间内实现了估值翻倍,显示出市场对其技术和潜力的极大认可。

💡 **颠覆性LPU技术**:Groq专注于打破英伟达在AI芯片领域的垄断,其核心产品并非传统的GPU,而是自称为“语言处理单元”(LPU)的推理引擎。这种专门优化的硬件旨在以极高的速度和效率运行AI模型,提供了一种区别于现有主流方案的创新路径。Groq的产品既可通过云服务提供,也可作为本地部署的硬件集群,满足不同客户的需求。

💰 **成本效益与开发者生态**:Groq宣称其产品在保持甚至提升AI性能的同时,能够显著降低运行成本。公司提供的服务支持Meta、DeepSeek、Qwen、Mistral、Google和OpenAI等主流模型的开源版本。目前,Groq已经服务了超过200万开发者,相较于一年前的35.6万开发者,用户增长迅猛,构建了一个日益壮大的开发者生态系统。

👨‍🔬 **深厚的技术背景**:Groq的创始人Jonathan Ross拥有与AI芯片开发紧密相关的行业经验。他曾参与谷歌TPU(Tensor Processing Unit)芯片的开发工作,而TPU是谷歌专门为机器学习任务设计的处理器,至今仍为谷歌云的AI服务提供支持。这一背景为Groq的技术创新和产品落地提供了坚实的基础。

AI chip startup Groq confirmed Wednesday that it raised a fresh $750 million in funding at a post-money valuation of $6.9 billion.  

This topped the rumored numbers when word leaked in July that Groq was raising. At that time, reports suggested that the raise would be about $600 million, at near a $6 billion valuation. 

Groq, which also sells data center computing power, previously raised $640 million at a $2.8 billion valuation in August 2024, making this more than double the valuation in about a year. Groq has now raised over $3 billion to date, PitchBook estimates.

Groq has been a hot commodity because it is working on breaking the chokehold that AI chip maker Nvidia has over the tech industry. Groq’s chips are not GPUs, the graphics processing units that typically power AI systems. Instead, Groq calls them LPUs, (language processing units) and calls its hardware an inference engine — specialized computers optimized for running AI models quickly and efficiently.  

Its products are geared toward developers and enterprises, available as either a cloud service or an on-premises hardware cluster. The on-prem hardware is a server rack outfitted with a stack of its integrated hardware/software nodes. Both the cloud and on-prem hardware run open versions of popular models, like those from Meta, DeepSeek, Qwen, Mistral, Google and OpenAI. Groq says its offerings maintain, or in some cases improve, AI performance at significantly less cost than alternatives.

Groq’s founder, Jonathan Ross, has a particularly relevant pedigree for this work. Ross previously worked at Google developing its Tensor Processing Unit chip, which are specialized processors designed for machine learning tasks. The TPU was announced in 2016, the same year Groq emerged from stealth. TPUs still power Google Cloud’s AI services.

Groq says it now powers the AI apps of more than 2 million developers, up from 356,000 developers when the company talked to TechCrunch a year ago.

Techcrunch event

San Francisco | October 27-29, 2025

The new round was led by investment firm Disruptive, with additional funding from BlackRock, Neuberger Berman, Deutsche Telekom Capital Partners, and others. Existing investors, including Samsung, Cisco, D1, and Altimeter, also joined the round.

Fish AI Reader

Fish AI Reader

AI辅助创作,多种专业模板,深度分析,高质量内容生成。从观点提取到深度思考,FishAI为您提供全方位的创作支持。新版本引入自定义参数,让您的创作更加个性化和精准。

FishAI

FishAI

鱼阅,AI 时代的下一个智能信息助手,助你摆脱信息焦虑

联系邮箱 441953276@qq.com

相关标签

Groq AI芯片 融资 估值 LPU 推理引擎 AI 科技初创 Nvidia Deep Learning Groq Funding AI Chip Valuation Inference Engine Startup
相关文章