For Immediate Release
June 11, 2025
Paris, France – Connect Tech, a leader in embedded Edge AI solutions, is proud to unveil its first NVIDIA Jetson Thor-powered AI demo at NVIDIA GTC Paris 2025. At Connect Tech’s booth (E24), this groundbreaking showcase highlights the power of NVIDIA’s next-generation platform in a real-world, deployable solution.
The debut coincides with Connect Tech’s 10-year milestone of supporting the Jetson ecosystem, underscoring a decade of delivering rugged, high-performance compute solutions for Edge AI deployments. A veteran in the embedded space, Connect Tech is also celebrating its 40th anniversary in June 2025.
“Being one of the first to showcase Jetson Thor in action is a major achievement,” said Rob Callaghan, Chief Product Officer at Connect Tech. “It demonstrates our continued commitment to Jetson developers and integrators, not just in concept, but in solutions they can deploy today.”
Live at Booth E24: Jetson Thor-Accelerated Video Search and Summarization (VSS)
The featured demo, Visual AI Agents on Jetson Thor: Video Search and Summarization (VSS), is a powerful generative AI application designed to simplify the development of intelligent video analytics agents. Based on the NVIDIA AI Blueprint for video search and summarization, VSS uses advanced vision-language models (VLMs), large language models (LLMs), computer vision, and VSS leverages Jetson Thor’s AI compute to run Agentic AI workloads at the Edge.
Demo Key Features
- Real-time and batch processing modes with 100 alerts processed in 10 seconds
- Video search across multiple live streams and recorded files
- Video summarization with contextualized insights
- Interactive Question and Answering (Q&A)
- Real-time alerts and notifications
- Object tracking with advanced computer vision
- Multimodal model fusion (VLM/CV/LLM integration)
- Audio support for comprehensive analysis
- Easy-to-use REST APIs for stream management and alert configuration
At the core of the VSS demo is Connect Tech’s Gauntlet Carrier Board integrated with the NVIDIA Jetson Thor system-on-module. This compact platform delivers the full performance of Jetson Thor in a deployable form factor ideal for real-world Edge environments. Gauntlet, built on Jetson Thor, provides the foundation to run advanced multimodal AI agents like VSS at full speed and efficiency.
Jetson Thor, based on NVIDIA’s Blackwell GPU architecture, delivers 2070 FP4 TFLOPS of AI compute, while maintaining exceptional energy efficiency at just 130W. This enables developers to deploy enterprise-grade generative AI models directly at the Edge without compromising accuracy or speed. The result is intelligent, responsive, and highly accurate video analysis systems that can process complex AI workloads locally, reducing latency, enhancing privacy, and enabling real-time decision-making in mission-critical applications.
Now Available: VSS as a Commercial Solution
Beyond the demo, Connect Tech is announcing that VSS will be available as a commercial offering. Utilize Connect Tech’s all-in-one hardware/software solution to integrate VSS into your organization’s video analytics workflows, accelerating development and deployment across security, industrial automation, smart cities, and more.
A Decade of Jetson Expertise
Connect Tech’s showcase at NVIDIA GTC Paris is the culmination of 10 years of dedication to the Jetson platform — from pioneering carrier boards to delivering complete ruggedized systems. Over the past decade, Connect Tech has remained at the forefront of Jetson innovation, supporting developers and integrators in building powerful AI-at-the-Edge applications.
About Connect Tech
Connect Tech is a global leader in embedded computing and Edge AI solutions for a wide range of industries, including industrial automation, public sector, robotics, transportation, and more. With a strong focus on innovation and customer satisfaction, Connect Tech offers a diverse portfolio of products designed to meet the demands of the most challenging embedded applications. Follow Connect Tech on LinkedIn, X, Facebook, Instagram, and YouTube.