Jetson Thor Delivers Edge AI Performance to Physical Systems

NVIDIA's Jetson Thor series packs up to 2,070 TFLOPS of FP4 AI performance, 128 GB of memory, high-speed sensors, and more.
Related To: 
Nov. 10, 2025
2 min read

What you'll learn:

  • Insight into NVIDIA's Jetson Thor platform and what it delivers in terms of AI performance, memory, and power for edge AI and robotics.
  • Why the high-core-count Arm CPU, Blackwell GPU, and advanced I/O suit the Jetson Thor for real-time physical-AI systems. 
  • How it can be utilized to scale embedded systems from multi-camera sensor fusion to generative-AI workflows at the edge.
William Wong/EndeavorB2B
69124a7b1c9b14b0b0872962 Bill Wong Jetson Thor

At the recent embedded world North America, NVIDIA introduced its latest embedded system module: the Jetson Thor series. Designed around the company's Blackwell GPU architecture, the Jetson Thor modules deliver up to 2,070 FP4 TFLOPS of AI compute, support up to 128 GB of LPDDR5X memory, and offer a configurable power curve from 40 to 130 W.

On the CPU side, the platform includes a 14-core Arm Neoverse-V3AE CPU, paired with a 2,560-core Blackwell GPU with 96 5th-Gen Tensor cores. Memory bandwidth is quoted at approximately 273 GB/s with a 256-bit LPDDR5X memory interface.

NVIDIA
nvidia_jetson_thor_alternative_promo_image

The Jetson Thor, in an AGX Developer Kit form, includes a 1-TB NVMe M.2 (Key M) slot, high-speed I/O including 4x 25 GbE networking, and carrier board features to support real-time multi-sensor fusion and robotics workflows.

The Jetson Thor marks a leap in computational and sensor-processing capabilities that enables physical AI workflows, including real-time inference, multi-camera arrays, active perception, generative AI models at the edge, and I/O integration. Thanks to the modular format, it can be scaled from prototype to production across robotics, automation, smart infrastructure, and other embedded systems.

>>Check out more of our embedded world North America 2025 coverage

ID 185716853 - Center © John Tolkovski - Dreadmstime.com
Embedded World North America 2025
Check out the latest technology and insights from the show.

About the Author

Cabe Atwell
Email

Cabe Atwell

Technology Editor, Electronic Design

Cabe is a Technology Editor for Electronic Design. 

Engineer, Machinist, Maker, Writer. A graduate Electrical Engineer actively plying his expertise in the industry and at his company, Gunhead. When not designing/building, he creates a steady torrent of projects and content in the media world. Many of his projects and articles are online at element14 & SolidSmack, industry-focused work at EETimes & EDN, and offbeat articles at Make Magazine. Currently, you can find him hosting webinars and contributing to Electronic Design and Machine Design.

Cabe is an electrical engineer, design consultant and author with 25 years’ experience. His most recent book is “Essential 555 IC: Design, Configure, and Create Clever Circuits”

Cabe writes the Engineering on Friday blog on Electronic Design. 

Sign up for our eNewsletters
Get the latest news and updates

Related

What’s the Difference Between DIMM and CAMM?
Taking the Data Center to the Next Level
DC-DC Converter Design Made Easy
Sponsored
MAX66250/MAX66301 NFC Secure Authenticators and Coprocessors
Sponsored

Voice Your Opinion!

To join the conversation, and become an exclusive member of Electronic Design, create an account today!