Hardware Wiki: GPU Wiki CPU Wiki

Tag MLPerf Inference

AI & Compute

AI & Compute Roundup: NVIDIA’s Vera Rubin Posts First MLPerf Numbers Next to AMD’s Biggest Cluster Yet

Bar chart, drawn to scale: NVIDIA Vera Rubin NVL72 inference throughput at 3.7x a GB300 NVL72 baseline on the Qwen3-VL MLPerf Inference v6.1 benchmark. Headline: MLPerf Inference v6.1, Vera Rubin debuts 3.7x faster.

NVIDIA's first Vera Rubin numbers land in MLPerf Inference v6.1 alongside AMD's biggest GPU cluster submission yet. Also this week: Apple's NVLink talks with NVIDIA, and a laptop with 160GB of usable VRAM.

Saeid Yazdani
September 17, 2026