NVIDIA Groq 3 LPX hits 3,431 tokens/s at 100K context
tomjaguarpaw · x · 2026-08-26
NVIDIA's technical blog details the Groq 3 LPX accelerator's performance on the 100K context benchmark. Paired with the Vera Rubin NVL72 platform, it achieved 3,431 output tokens/second on the Gemma 4 31B model, nearly 4x faster than the next fastest public endpoint. The system utilizes deterministic compiler-scheduled workload planning and fine-grained computation-communication overlap to minimize latency, enabling high-interactivity, long-context inference for multi-agent systems.
More from Infra
- Jensen Huang joins BG2 podcast: does AI revenue justify capex, '27 bottlenecks — elonmusk · 2026-10-16
- Mistral's on-device infra rack revealed: DGX Spark, Mac Studio and more — gregmushen · 2026-10-16
- NVIDIA and Sweden's Wallenberg launch Sferical One, a shared AI supercomputer — nvidia · 2026-10-16
- Next infra frontier: self-evolving systems that adapt and self-verify — _shreya_s · 2026-10-16
- Xiaomi's MiMo RL dashboard reveals $3.47M spent on just 30 training steps — bycloud · 2026-10-16
- Cerebras pushes disaggregated inference: heterogeneous prefill pools feeding wafer-scale decode — ai · 2026-10-16