Dell Partners with Groq and NVIDIA on Enterprise AI Inference

On August 25, Dell CEO Michael Dell announced a partnership with Groq and NVIDIA to bring a new enterprise AI inference solution to market. Groq will be among the first adopters of the NVIDIA Groq 3 LPX accelerator, deploying it alongside the NVIDIA Vera Rubin NVL72 platform in Groq's own AI inference cloud, with Dell handling hardware integration.

Confirmed

• The solution pairs the NVIDIA Groq 3 LPX accelerator with the Vera Rubin NVL72 platform in a dedicated AI inference cloud
• Dell's benchmarks show output speed of 3,400 tokens per second, with 4x higher interactivity than recent competitors
• According to @rohanpaulai, this marks the first appearance of Groq technology in an NVIDIA rack-scale product, 8 months after their non-exclusive licensing deal, with Groq racks going live this year
• Groq confirmed on its official account that it will deploy the hardware with Dell
• NVIDIA is splitting agentic AI workloads across specialized processors: Rubin GPUs handle heavy model compute while other tasks go to Groq chips

Why it matters

• This is the first time Groq's chip technology enters an NVIDIA rack-scale product, marking the tangible landing of their licensing collaboration
• Splitting inference compute by workload offers a new hardware architecture option for agentic AI, and Dell's involvement accelerates enterprise commercialization

2026-08-25 ~ 2026-08-25 · 5 related posts

• [source] Dell partners with Groq for new AI inference class hitting 3,400 tokens/sec — Dell · 2026-08-25
• Dell partners with Groq and Nvidia to deploy next-gen AI inference cloud — IanAndrewsDC · 2026-08-25
• [source] NVIDIA splits agentic AI workloads across Rubin GPUs, Groq LPX and Vera CPUs — rohanpaul_ai · 2026-08-25
• Dell Partners with NVIDIA and Groq to Deploy 3,400 Tokens/sec Inference Solution — IanAndrewsDC · 2026-08-25
• [source] Groq to Adopt NVIDIA Groq 3 LPX Accelerator — GroqLLC · 2026-08-25