• Powered by 4x Hailo-8™ AI inference processors
• Up to 104 TOPS AI performance
• Typical power consumption:25W
• Provides comprehensive software package such as AI development tools
• Supports fine-tuning customization service for AI inference models integration
AI-MXM-H84A features the first-ever MXM 3.1 Type B module powered by four Hailo-8™ AI inference processors, providing up to 104TOPS AI performance with high power efficiency and multitasking to speed up deployment of neural network (NN) and deep learning (DL) processes on edge devices.
Because of size, weight and power(SWaP) reduction, the AI accelerator can be easily integrated into a variety of embedded systems to handle heavy inference workloads with low latency. For example, AI-MXM-H84 is widely applied to automatic guided vehicles (AGV), virtual fencing system, autonomous machines and computer vision systems.
Users can shorten the AI projects development cycles via comprehensive software package such as AI development tools and customization services, including industrial-grade mechanism design, fine-tuning for AI inference models integration.
|Engine Specs||4x Hailo-8 AI processor with up to 26 TOPS and best-in-class power efficiency|
|AI Performance||104 TOPS|
|Feature Support||PCI Express 3.0 x16
Supported TensorFlow and ONNX
|Power Consumption||25 W (Typical power consumption)|
|Form Factor||MXM graphics module version 3.1, Type B|
|Dimensions (WxD)||82.0 x 105.0 mm (3.22” x 4.13”)|
|Net Weight||0.05 kg (0.1102 lb)|
|Vibration||2.4Grms @5~500 Hz, Sine, 0.5Hr/axis|
|Temperature||Standard: Operating Temp. : 0 to + 70°C (32°F ~158°F )/
Storage Temperature: -40 to + 85°C (-40°F ~ 185°F)
|Humidity||90% @ 40°C Related Humidity, Non-condensing|
|OS Support||Windows 10/11 64-bit
MXM3.1 Type B, 4x Hailo-8, 0°C to +70°C