Groq to Adopt NVIDIA Groq 3 LPX Accelerator
GroqLLC · x · 2026-08-25
Groq announced it will be among the first adopters of the NVIDIA Groq 3 LPX accelerator, deploying it alongside the NVIDIA Vera Rubin NVL72 platform in its AI inference cloud. Groq is working with Dell for the deployment. NVIDIA's technical blog indicates Groq 3 LPX achieved 3,431 output tokens/second with the Gemma 4 31B model and supports various disaggregation configurations for multi-trillion parameter models.
Related event: Dell Partners with Groq and NVIDIA for Next-Gen AI Inference(4 posts)→
More from Infra
- Supabase MCP server now supports enterprise-managed auth via Okta — dshukertjr · 2026-08-25
- LocalAI Checker tool detects hardware compatibility for local models — dr_cintas · 2026-08-25
- Cerebras/Groq chaining thousands of chips enables 100T scaling — scaling01 · 2026-08-25
- Livestream: Running ComfyUI Locally via MCP and Hardware Optimization — MiniMax_AI · 2026-08-25
- Vercel AI Gateway charges zero markup and offers discounts — cramforce · 2026-08-25
- Should progressives be data centers' biggest fans given their jobs and footprint? — pmddomingos · 2026-08-25