EDGE AI LAB INFRASTRUCTURE REQUIREMENT
DOCUMENT
1. INTRODUCTION
The proposed Edge AI Laboratory is designed to provide a comprehensive platform for AI education,
research, prototyping, robotics development, computer vision applications, autonomous systems, industrial
automation, IoT analytics, and edge computing workloads. The laboratory consists of NVIDIA Jetson-based
development platforms, NVIDIA AGX Orin systems, NVIDIA Jetson Xavier systems, NVIDIA IGX Thor
platform, AI Robotic Car Kits, Seed Studio Edge AI systems, and Lanner Edge AI systems.
Unlike centralized GPU servers, Edge AI devices perform AI inferencing and application processing directly
at the edge, enabling real-time decision-making with low latency and reduced dependency on cloud
infrastructure. The proposed lab environment allows students, researchers, faculty members, and
developers to build, train, optimize, deploy, and validate AI models across multiple hardware platforms
while gaining hands-on experience with industry-standard edge computing technologies.
The proposed Edge AI Lab includes approximately 66 Edge AI devices and development platforms that can
be simultaneously utilized for educational activities, project development, AI model validation, robotics
experimentation, industrial AI research, and collaborative innovation programs.
2. PROPOSED EDGE AI LAB EQUIPMENT
The proposed laboratory consists of the following systems:
• NVIDIA Jetson AGX Xavier 64GB – 2 Units
• NVIDIA AGX Orin Kits – 8 Units
• NVIDIA Jetson Orin Nano Developer Kit 8GB – 10 Units
• NVIDIA Jetson Nano (B01) 4GB – 25 Units
• NVIDIA IGX Thor Platform – 1 Unit
• AI Robotic Car Kits – 5 Units
• Seed Studio Recomputer J4012 with NVIDIA Jetson Orin NX 16GB – 10 Units
• Lanner EAI-I130 AI Edge System with NVIDIA Jetson Nano – 5 Units
Total Edge AI Systems: Approximately 66 Units
1
3. NETWORK INFRASTRUCTURE REQUIREMENT
A robust network infrastructure is essential to support AI model deployment, software package updates,
remote administration, centralized storage access, and collaborative research activities.
Local Area Network (LAN)
A minimum 1 Gigabit Ethernet network is recommended for the entire laboratory. All Edge AI devices,
workstations, storage systems, and management systems should be connected through managed Gigabit
Ethernet switches to ensure reliable communication and high-speed data transfer.
Recommended Network Components:
• Managed Layer-2 Gigabit Ethernet Switch
• Enterprise Router/Firewall
• Structured CAT6 Cabling
• Dedicated VLAN for AI Laboratory
• Centralized Network Management
• Network Access Control Policies
Internet Connectivity
Internet connectivity is primarily required for:
• NVIDIA SDK Downloads
• CUDA Toolkit Updates
• JetPack Software Installation
• AI Framework Downloads
• Dataset Downloads
• Cloud-Based AI Services
• Remote Collaboration
• Software License Activation
2
Recommended Internet Bandwidth:
Basic Academic Usage: 100 Mbps Dedicated Internet
AI Development and Research: 300 Mbps Dedicated Internet
Multi-User AI Training Environment: 500 Mbps
Advanced Research Laboratory: 1 Gbps Internet Connectivity
Wireless Connectivity
Enterprise-grade Wi-Fi 6 infrastructure is recommended throughout the laboratory.
Recommended Wi-Fi Features:
• Wi-Fi 6 Access Points
• Dual Band Operation
• High Device Density Support
• Secure Authentication
• Separate Student and Faculty Networks
• High-Speed Wireless Connectivity for Laptops and Mobile Devices
4. POWER INFRASTRUCTURE REQUIREMENT
One of the major advantages of an Edge AI Laboratory is its significantly lower power consumption
compared to traditional AI data center infrastructure.
The estimated power consumption for the proposed laboratory is as follows:
Jetson AGX Xavier Systems: 60–90 Watts
AGX Orin Systems: 200–320 Watts
Orin Nano Systems: 100–150 Watts
Jetson Nano Systems: 175–250 Watts
NVIDIA IGX Thor Platform: 60–100 Watts
3
AI Robotic Car Kits: 100–150 Watts
Seed Studio Edge AI Systems: 150–250 Watts
Lanner Edge AI Systems: 50–100 Watts
Total Estimated Laboratory Load:
Approximately 3 kW to 5 kW under full utilization conditions.
Recommended Electrical Infrastructure:
• 230V AC Single Phase Supply
• Dedicated Laboratory Power Circuit
• Proper Earthing and Grounding
• MCB Protection
• Surge Protection Devices
• Industrial Grade Power Distribution Units
• Electrical Safety Compliance
Recommended UPS Backup:
Minimum UPS Capacity: 5 kVA
Preferred UPS Capacity: 10 kVA
UPS backup ensures uninterrupted operation during power fluctuations, short outages, software upgrades,
and ongoing experiments.