Groq Raises $750M Series E at $6.9B Valuation to Scale AI Inference Hardware

San Jose, CA — May 28, 2026 — Groq, developer of the Language Processing Unit (LPU) chip architecture, has closed a $750 million Series E funding round at a $6.9 billion post-money valuation, bringing total capital raised to over $1.5 billion since the company's 2016 founding. The round positions Groq to accelerate LPU production and expand enterprise deployment of its specialized AI inference infrastructure.

The valuation represents a 146% increase from Groq's estimated $2.8 billion Series D valuation in 2024, reflecting sustained investor confidence in purpose-built inference hardware as enterprises shift AI workloads from training to production deployment at scale.

"Inference is where AI meets reality," said Jonathan Ross, Groq CEO and Founder. "Every token generated, every query answered, every agent decision made runs on inference hardware. We built the LPU specifically for this moment—and this funding allows us to meet demand that is growing faster than the industry anticipated."

Addressing the Inference Bottleneck

As large language model deployments have expanded across enterprise environments, inference latency and power consumption have emerged as dominant cost variables in AI infrastructure budgets. Groq's LPU architecture—designed by Ross, a former Google TPU architect—targets this specific constraint through a linear processing model that reduces memory bandwidth requirements compared to conventional GPU approaches.

Groq's published benchmarks cite inference speeds of up to 500 tokens per second for LLM workloads, claiming 10–100x latency improvements over GPU alternatives in specific use cases, alongside 3–5x power efficiency gains. For hyperscalers and enterprises operating inference at scale, those figures translate directly to data center economics and total cost of ownership.

The global AI inference chip market is projected to reach $45–60 billion by 2026, growing at a 35–45% compound annual rate, according to estimates from IDC and Gartner. Groq is targeting an addressable segment of $15–25 billion within that market.

Enterprise Traction and Cloud Integration

Groq reported more than 100 enterprise customers in pilot or production deployments as of 2026, including Fortune 500 companies across financial services, healthcare, and technology sectors. Early customers have reported total cost of ownership reductions of 50–70% versus incumbent GPU-based inference infrastructure, according to company statements.

Strategic integrations with major cloud providers—including AWS, Google Cloud, and Microsoft Azure—have expanded Groq's distribution reach beyond direct enterprise sales. These partnerships provide access to managed inference services and position Groq within existing procurement workflows for enterprise AI teams.

Competitive Landscape

NVIDIA currently holds an estimated 85–90% share of the AI chip market, with its H100 and H200 GPU lines serving as the default infrastructure for both training and inference workloads. Groq and a cohort of specialized competitors—including Cerebras, SambaNova, and cloud-native solutions such as AWS Inferentia and Google TPU—are collectively targeting the remaining market share and future growth segments.

NVIDIA's forthcoming Blackwell and Rubin architectures include inference-specific optimizations. Groq has indicated it intends to maintain its performance lead through continuous LPU iteration, with next-generation chip development funded in part by the Series E proceeds.

Capital allocation from the round is expected to cover expanded TSMC manufacturing capacity, developer tooling and SDK investment, enterprise sales team growth, and research and development for subsequent LPU generations.

About Groq

Groq was founded in 2016 by Jonathan Ross, former architect of Google's Tensor Processing Unit. Headquartered in San Jose, California, the company designs and manufactures the Language Processing Unit (LPU), a purpose-built inference processor optimized for large language model and generative AI workloads. Groq's hardware and cloud inference platform serve enterprise customers across financial services, healthcare, and technology sectors, with integrations across major public cloud providers. The company has raised over $1.5 billion in total funding.

Media Contact: [email protected] | groq.com