---
format: "aidr-story-markdown/v1"
id: "481380da69e9d0c104790420f6d0dfd7067a93b90903f3cc40cdc01b1ff66738"
canonical_url: "https://aidr.today/481380da?lang=en"
title: "vLLM Hits 7.8x GB200 Throughput With NVIDIA Vera Rubin Support"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-10-10T02:39:15.000Z"
category: "Open Source"
topics: ["nvidia","inference","open-source","llm","framework"]
source_urls: ["https://marketbrief.now/ai/vllm-hits-78x-gb200-throughput-with-nvidia-vera-rubin-support-cd147fec","https://huggingnews.com/ai/vllm-hits-78x-gb200-throughput-with-nvidia-vera-rubin-support-cd147fec"]
summary: "vLLM has added compatibility for NVIDIA's Vera Rubin GPU architecture to its open-source inference library. Early tests using the MiniMax M3 model on AgentX show more than 7.8x the throughput of the GB200."
---

# vLLM Hits 7\.8x GB200 Throughput With NVIDIA Vera Rubin Support

> [Open the canonical story](<https://aidr.today/481380da?lang=en>)

**Published:** 2026-10-10T02:39:15.000Z
**Category:** Open Source
**Topics:** nvidia, inference, open\-source, llm, framework

## Summary

vLLM has added compatibility for NVIDIA's Vera Rubin GPU architecture to its open\-source inference library\. Early tests using the MiniMax M3 model on AgentX show more than 7\.8x the throughput of the GB200\.

## Sources

- [Story source](<https://marketbrief.now/ai/vllm-hits-78x-gb200-throughput-with-nvidia-vera-rubin-support-cd147fec>)
- [Story source](<https://huggingnews.com/ai/vllm-hits-78x-gb200-throughput-with-nvidia-vera-rubin-support-cd147fec>)