GW16168-SP520
GW16168-SP520
Recommended for New Design (RND)
The GW16168-SP520 is an 8GB memory variant of the USA-made GW16168 M.2 AI accelerator. Powered by the NXP Ara240 DNPU, it delivers up to 40 eTOPS of dedicated edge AI performance for machine vision, robotics, automation, and embedded AI inference.
Specifications
Features
Additional Description
M.2 edge AI accelerator for industrial embedded systems and real-time AI inference
Security
Secure Boot and Root of Trust
Additional Features
Up to 40 eTOPS of dedicated edge AI performance
Software
Software
TensorFlow, PyTorch, and ONNX support
Packaging
Form Factor
M.2 M-Key 2280
Dimensions
22.0 x 80 x 3.5 mm
Interface
Included Peripheral
8 GB onboard LPDDR4 DRAM
Physical Interface
PCIe Gen4 x4; compatible with PCIe Gen3 hosts
Power
Power Consumption
2.0 A typical operating current at 3.3 V
Input Voltage
3.3 V M.2 edge connector
Processing
Chipset
NXP Ara240 Discrete Neural Processing Unit (DNPU)
Environment
Temperature Range
-40°C to +85°C
