# CUDA 护城河警报：CUDA vLLM 支持 DeepSeek v4.1 Flash 两天后，AMD 才发布镜像

- 来源：SemiAnalysis (@SemiAnalysis_)
- 发布时间：2026-09-12 11:45
- AIHOT 分数：49
- AIHOT 链接：https://aihot.news/items/cmtxvctm305ykrous9zb47ruq
- 原文链接：https://x.com/SemiAnalysis_/status/2098618867035557984

## AI 摘要

CUDA vLLM 支持 DeepSeek v4.1 Flash 两天后，AMD 才公开发布其 DeepSeek v4.1 Flash 镜像。该镜像功能上开箱即用，但每美元性能目前比 H200 差最多 14.8 倍、比 B200/B300 差最多 42 倍。

## 正文

POWER OF CUDA MOAT ALERT🚨: 2 days after CUDA vLLM supported DeepSeekv4.1 Flash, AMD finally publicly released its DeepSeek v4.1 Flash image. Functionally, it works out of the box, but performance-wise, it is currently up to 14.8x worse perf per dollar than H200 and up to 42x worse perf per dollar than B200/B300 currently.

The 🚀 POWER OF THE CUDA MOAT 🚀 is that NVIDIA's collaboration with its massive 6 million-developer community ecosystem means that CUDA is optimized on day 0. As AMD Anush said, "Speed is the Moat," and day 0 model support shows CUDA is the speed.