YUAN Unveils Pandora: Ready-to-Deploy Edge AI Designed for Developers
Weighing just 470g, Pandora delivers 157 TOPS of AI performance and offers modular flexibility tailored to developers' needs.
YUAN High-Tech, a leader in AI and imaging solutions, announces Pandora, an ultra-compact edge AI platform powered by NVIDIA® Jetson Orinâ„¢ NX Super mode. Already deployed in education, smart retail, and robotics, Pandora enables real-time analytics, localized AI-driven services, and autonomous systems-without relying on the cloud.
Weighing just 470g, Pandora delivers 157 TOPS of AI performance and offers modular flexibility tailored to developers' needs.
Key Features:
Dual Gigabit Ethernet ports
4× USB ports (with OTG support)
HDMI 2.0 and audio I/O
4× M.2 slots (for SSD, Wi-Fi, Bluetooth, 5G/LTE)
MIPI CSI, UART, CAN Bus interfacesyu
GPIO header for industrial/IoT integration
Optional HDMI/SDI capture card
Detachable housing for 3D-printed customization
Pandora supports the full NVIDIA Jetson ecosystem, including Jetson Platform Services, TAO Toolkit, and Metropolis Microservices. It runs mainstream AI models such as LLaMA, ChatGLM, Stable Diffusion, and ViT, enabling NLP, CV, OCR, and generative AI tasks. YUAN's proprietary modules further support real-time object, face, and behavior recognition.
About YUAN High-Tech
YUAN delivers advanced video capture and edge AI solutions for industrial and commercial use.
Featured Product
Strain Wave Gearheads by maxon - Highly precise, compact, and efficient.
Strain wave gearheads are ideal for use in applications requiring precision motion control and high torque transmission in a compact design. maxon strain wave gearheads are a specialized type of gearhead where precision and compact design are crucial. This type of gearhead is often used in applications in a confined space. With their advanced gear tooth design, maxon's strain wave gearheads provide high efficiency and good back drivability. Typical applications include: Robotics, Aerospace, Medical technology and Optical systems .
