Groq announced a $640 million Series D at a $2.8 billion valuation for LPU inference, the company announced on 5 August 2024 via PR Newswire. The round was led by BlackRock Private Equity Partners, with Cisco Investments and others participating.
Inference specialists are raising large rounds as enterprises look past training hype to tokens-per-second and cost. Buyers still need workload-specific benchmarks and clear multi-cloud deployment options.
Groq has pitched Language Processing Units for deterministic, low-latency serving. The Series D is fuel to meet demand as model builders and enterprises shop for alternatives to GPU-only stacks.
Infrastructure buyers should insist on reproducible latency and price tests on their models — before treating any LPU funding headline as a completed inference strategy.