← Back to search

Did Frontier AI Just Move to the Edge?

JetsonHacks 10:02

1,688 views · 84 likes Watch on YouTube ↗

What happens when yesterday’s frontier-class AI capabiltiy starts fitting on edge hardware?

In this video, we run Qwen 3.8 27B and Nemotron 3.5 Lightning on Jetson AGX Thor, compare their very different capability and performance profiles, and look at how NVIDIA Switchyard can route work between models. Also, does Blackwell introduce a major advantage with native NVFP4?

00:00 Introduction
02:10 Qwen 3.8 27B Benchmarks
04:45 Nemotron 3.5 Lightning Benchmarks
05:10 Speculative Decoding
06:04 Qwen vs Nemotron Lightning IRL
08:50 NVIDIA Switchyard - Policy Router
09:11 Deep Insights

Jetson AI Lab Models: https://www.jetson-ai-lab.com/models/
NVIDIA Switchyard: https://github.com/NVIDIA-NeMo/Switchyard
llama-benchy: https://github.com/eugr/llama-benchy
Artificial Analysis: https://artificialanalysis.ai

Join this channel to get access to perks:
https://www.youtube.com/channel/UCQs0lwV6E4p7LQaGJ6fgy5Q/join

As an Amazon Associate I earn from qualifying purchases.
Visit the JetsonHacks storefront on Amazon: https://www.amazon.com/shop/jetsonhacks

Visit the website at https://jetsonhacks.com
Sign up for the newsletter! https://newsletter.jetsonhacks.com
Github accounts: https://github.com/jetsonhacks
https://github.com/jetsonhacksnano
Twitter: http://twitter.com/jetsonhacks

Some of these links here are affiliate links. As an Amazon Associate I earn from qualifying purchases at no extra cost to you.

Category (YouTube): Science & Technology

Playback is via YouTube's official embedded player. Data from YouTube; Exumo is not affiliated with YouTube.