Groq

Latest AI news, models and releases from Groq. ['Groq 3 LP30', 'Groq 3 LPU', 'Groq 3 LPX', 'llama-3.3-70b', 'LPU', 'RealScale']

Official site ↗

Hardware & Inference 🇷🇺 ⚡

NVIDIA Groq 3: SRAM, Disaggregation and Determinism in the New AI Platform

NVIDIA's $20 billion acquisition of Groq has resulted in the integration of LPU accelerators into the Vera Rubin platform, enabling disaggregated, heterogeneous AI inference. The new Groq 3 LPX rack system combines 256 LPUs for fast, deterministic token generation, while Vera Rubin NVL72 handles flexible training and inference. The platform targets low-latency interactive AI and multi-agent workloads, promising up to 35x faster performance than Grace Blackwell NVL72 at 400 TPS per user.

NVIDIA

Groq

ServerNews27.07 · 15:06

Hardware & Inference 🇷🇺 ⚡

NVIDIA Unveils Vera Rubin POD Cluster: Five Racks, Seven Chips, 1.2 Quadrillion Transistors

NVIDIA has introduced the Vera Rubin POD, an AI cluster built on third-generation MGX architecture, comprising five specialized rack platforms with seven chip types. The system integrates 40 racks, 1.2 quadrillion transistors, and 1152 Rubin GPUs, delivering 60 EFLOPS performance and 10 PB/s interconnect speed. The cluster includes NVL72, LPX, Vera, STX, and SPX racks, targeting agentic AI workloads.

NVIDIA

Groq

ServerNews27.07 · 15:05

Show more ↓