產品特色

• Intel® Core™ Ultra processors feature CPU/GPU/NPU hybrid AI engines, enhancing CPU productivity by up to 14%
• Intel® Core™ Ultra processors integrated with maximum 34 TOPS
• Multiple HDMI and DisplayPort display interfaces support max 5 independent displays, up to 4K resolution
• 2 USB 3.2 Gen 2, 1 USB 3.2 Gen 2x2 Type-C with DP, 2 USB 2.0, 1 2.5GigE LAN with Intel® TSN
• 3 M.2 Socket for storage and expansion
• Modular Design for flexible expansion: USB, Isolated DIO, COM, LAN or 4G/LTE
• DC 12V to 24V Power Input, 0°C to 55°C operation
• Optional VHub One-Stop AIoT Solution Service supports OpenVINO™ toolkit for AI Computing

 

產品亮點

Optimized AI Platform

• Intel Core Ultra Processor features performance hybrid architecture with new AI experiences, immersive graphics, and an optimal balance of battery life and performance, delivers max 14% CPU productivity improved and up to 34 TOPS. 
• Multiple compute engines in one SoC : CPU (6 P-cores, 8 E-cores, 2 Low Power E-cores), GPU (Intel Arc GPU), and NPU (Intel AI Boost).
• High-speed DDR5 Memory, up to 96GB
• Intel Arc graphics supports DirectX 12.2, OpenGL 4.6, OpenCL 3.0, and Intel® Deep Learning Boost.
• High-speed data transfer : USB 3.2 Gen 2x2 Type-C (20G), PCIe 4.0 (16GT/s), USB 3.2 Gen 2 (10G), 2.5GigE (2.5G) 



 

Compact Integration

• Small Form Factor, compact size
• DC 12V to 24V Wide Range Power Input
• Multiple HDMI and DisplayPort, max 3 independent displays, up to 4K resolution
• 2.5GigE LAN, 10G USB 3.2 Gen 2, USB 2.0
• Supports Compact M.2 Key M Storage and M.2 Key E Expansion
• Smart Manageability : Intel vPro, TCC, TSN, PXE, Wake on LAN
• EN50155, EN50121-3-2
• OpenVINO toolkits support 500+ AI models optimized for AI Computing



 

AI Accelerating Options  

• Optional supports Compact MXM Graphics to boost performance AI and media processing workloads   
• NVIDIA Ada Lovelace/Ampere/Turing architecture supports max 9728 NVIDIA CUDA cores, 384 Tensor Cores, or 76 RT Cores, delivering max 41.15 TFLOPS peak FP32 performance
• Intel Xe HPG microarchitecture delivers 72 TOPS AI performance 
• Compact MXM Accelerator : NVIDIA RTX A2000 Ada, NVIDIA RTX A2000, NVIDIA Quadro T1000, NVIDIA Quadro RTX 3000, NVIDIA Quadro RTX 5000, NVIDIA Quadro RTX A3500 Ada, NVIDIA Quadro RTX 5000 Ada, Intel Arc A370M



 

Versatile Configurations 

• Innovative Stackable Docking design for optional expansion functions
• TGS-101 : 16-bit GPIO
• TGS-102 : 16-bit Isolated DIO (8 DI, 8 DO)
• TGS-104 : 2 Isolated COM RS-232/422/485 
• TGS-105 : Dual USB 3.0
• TGS-106 : Dual 1G LAN
• TGS-107 : 4G Module with SIM Socket