# NVIDIA Vera Rubin NVL72 Delivers Leading Performance in MLPerf Inference v6.1 Debut — retained record

> Retained reddit record from reddit:NVDA_Stock.

Canonical: https://highsignal.app/data/records/c6449b1f07c45155

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:NVDA_Stock)  published 2026-09-16 16:31 ingested 2026-09-17 02:38

[original source ↗](https://www.reddit.com/r/NVDA_Stock/comments/1wi27b5/nvidia_vera_rubin_nvl72_delivers_leading/)

## retained text



```
The latest MLPerf Inference benchmark tests were just released.  Results are released quarterly, alternating between inference and training. 

Looks like Nvidia is firing on all cylinders - Vera Rubin is meeting performance expectations and volume deliveries are underway.

From the blog post:

- In its first MLPerf Inference preview submission, NVIDIA Vera Rubin NVL72 delivers up to 3.7x better throughput than GB300 NVL72.

- NVIDIA GB300 NVL72 scales with leading efficiency: A 288-GPU submission across four GB300 NVL72 racks achieved 99% scaling efficiency, with throughput growing nearly linearly from a single-rack baseline.

- Continuous software optimizations drive performance gains: Software optimizations in NVIDIA’s MLPerf Inference v6.1 submissions delivered up to 1.6x higher performance over v6.0. Optimizations continued post-v6.1 submission, delivering further performance gains.
```



document fetched 2026-09-17 00:36 · canonical https://reddit.com/r/NVDA_Stock/comments/1wi27b5/nvidia_vera_rubin_nvl72_delivers_leading
