# @nvidia on YouTube

- **Type:** Video
- **Original URL:** https://youtube.com/watch?v=yYha_OtxA14
- **Gondola URL:** https://gondola.cc/posts/35554093-nvidia-youtube
- **Thumbnail:** https://img.gondola.cc/tr:w-,h-,fo-auto/postThumbnails/e3ee82ed20.jpg
- **Posted:** 2025-10-09T23:43:25.000+00:00
- **Account Owner:** NVIDIA (@nvidia) — https://gondola.cc/nvidia

## Caption

Industries worldwide are scaling AI as never before, and NVIDIA leads the way with unmatched #inference performance at #AIfactory scale. The NVIDIA Blackwell platform is the standard for next-generation AI, as demonstrated by independent open source results from SemiAnalysis InferenceMAX:

DeepSeek-R1: One GB200 NVL72 system can generate $75M in token revenue in three years, delivering 15x ROI for AI factories.

gpt-oss-120B: Leveraging TensorRT-LLM, B200, and the NVLink Switch’s 1,800 GB/s bandwidth, NVIDIA achieves 60,000 tokens/sec per GPU and 1,000 tokens/sec real-time per user.

Llama 3.3 70B: For dense open models, NVIDIA Blackwell delivers 10,000 tokens/sec per GPU—a 4x performance boost compared to Hopper.

With 10× greater performance per watt compared to Hopper, organizations slash costs and achieve breakthrough energy efficiency.

Explore how NVIDIA’s full-stack innovation unlocks new levels of efficiency, scalability, and profitability—whether in healthcare, manufacturing, the public sector, or beyond.

📝 See key results: https://nvda.ws/4nu9W3n 
🛠️ Explore technical details: https://nvda.ws/48t4lpv

## Stats

- **Views:** 34,448
- **Likes:** 764
- **Shares:** 0
- **Comments:** 0

## Tags

aifactory, inference

---
Copyright (c) Gondola