HAILO-8™ AI Module

Unprecedented AI Performance for Edge Devices

Top Performing Edge AI Accelerators

Hailo offers high performance edge AI processor, with a superior area and power efficiency.

At a size smaller than a penny even including the required memory, and with an architecture that takes advantage of the core properties of neural networks, our processors allow edge devices to run deep learning applications at full scale more efficiently, effectively, and sustainably than other AI chips and solutions, while significantly lowering costs.

Overview　|　Module Details　|　Tech Spec　|　Quick Start Guide & Downloads

Overview

Powerful

High performance – powered by Hailo-8™ AI processor with 26 Tera Operations Per Second (TOPS)
Enabling real-time, low latency, and high-efficiency AI inferencing on edge devices
Robust software suite supports state-of-the-art deep learning models & applications out-of-the-box

Efficient

Best-in-class power efficiency
Highest cost-efficiency (TOPS/$) compared with existing solutions

Scalable

Enabling simultaneous processing of multi-streams & multi-models
* Supporting extended temperature range of -40°C to 85°C

Module Details

Introduction to Developing and Deploying AI Applications with Hailo

Learn about the hardware and software that is helping to streamline the development and deployment of AI applications.

For users seeking to accelerate time to market, these best-in-class tools from Hailo and UP Bridge the Gap will reduce development time and effort, while reaching unprecedented levels of AI performance for edge devices.

Hailo 8 Getting Started Guide with UP compatible boards (based on Tappas 3.4

🔸 Curious how an AI module delivers 26 TOPS at just 2.5 watts?

🔸 Struggling about your AI development progress?

🔸 Looking for a sample application that can speed up your project?

Hailo-8 M.2 Module offers a breakthrough microprocessor uniquely designed to accelerate embedded AI applications on edge devices.

Compatible Products

▶ Add UP Squared Pro
1x Hailo-8™ on UP Squared Pro

　▶ Add UP Squared 6000
1x Hailo-8™ on UP Squared 6000

　▶ Add UP Xtreme i11
1x Hailo-8™ on UP Xtreme

Supported UP Products
The Hailo-8 M.2 2280 B+M key AI Acceleration module is compatible with UP Squared Pro, UP Squared 6000, and UP Xtreme i11 which are powerful computing platforms built for industrial applications to scale the Industry Edge Solution.

UP Squared Pro features the Intel Atom® Series processors (formerly Apollo Lake), which run at speeds up to 2.5 GHz.
UP Squared 6000 comes with the latest Intel Atom® Series processors (formerly Elkhart Lake) which offer up to 1.7x better performance and twice the performance for 3D graphics compared with the previous generation UP Squared Pro (formerly Apollo Lake).
UP Xtreme i11 is powered by the latest 11th Generation Intel® Core™ Processors, which can clock up to 4.4 GHz with only 28 watts TDP and 15 watts cTDP. For the even more power-efficient performance at the edge.

UP Products	UP Squared Pro	UP Squared 6000	UP Xtreme i11
Dimension	101.6mm x 101.6mm	101.6mm x 101.6mm	120 mm x 122mm
CPU	Intel® Celeron® N3350 (up to 2.4 GHz) Intel® Pentium® N4200 (up to 2.5 GHz) Intel Atom® x7-E3950 (up to 2.0 GHz)	Intel® Celeron® N6210 (up to 2.6 GHz) Intel® Pentium® J6426 (up to 3.0 GHz) Intel Atom® x6413E (up to 3.0 GHz) Intel Atom® x6425RE (up to 1.9 GHz)	Intel® Core™ i7-1185GRE (up to 4.4 GHz) Intel® Core™ i5-1145GRE (up to 4.1 GHz) Intel® Core™ i3-1115GRE (up to 3.9 GHz) Intel® Celeron® 6305E (up to 1.8 GHz)
RAM	2GB, 4GB, 8GB LPDDR4	2GB, 4GB, 8GB LPDDR4	2x DDR4 2 channel SO-DIMM VERTICAL MAX 64GB @ 3200MT/s
Storage	eMMC 32 GB / 64 GB	eMMC 32 GB / 64 GB	via 1x SATA 3.0 connector with power connector
Expansion slot used for Hailo module	1x M.2 M key 2280 (PCIe Gen-2.0, 1 lane)	1x M.2 M key 2280 (PCIe Gen-3.0, 2 lanes)	1x M.2 M key 2280 (PCIe Gen-3.0, 2 lanes)
Supported OS	Ubuntu 20.04 host OS, Hailo software via Ubuntu container	Ubuntu 20.04 host OS, Hailo software via Ubuntu container	Ubuntu 20.04 host OS, Hailo software via Ubuntu container

This table is also available on our wiki.

Delivering Data-Center Class Performance to Edge Devices

The Hailo-8™ unique, powerful and scalable structure-driven dataflow architecture takes advantage of the core properties of neural networks. It enables edge devices to run deep learning applications at full scale more efficiently, effectively, and substantially than traditional solutions, while significantly lowering costs.

Tech Specifications

Hailo-8™ AI
Form factor	M.2 key B+M
Interface	PCIe Gen-3.0, 2-lanes
Supported OS	Linux, Windows
Dimensions	Key B+M: 22×42 mm with breakable extensions to 22×60 mm & 22×80 mm
Supported host architectures	X86 or ARM based
Supported AI frameworks	TensorFlow, TensorFlow Lite, Keras, PyTorch & ONNX

AI Development Kit

UP Squared AI Vision Development Kit –

Implement and deploy computer vision solutions and support deeplearning inference across the CPU, GPU, and VPU.

This kit includes UP Squared x7-E3950 CPU 8GB RAM, 64GB eMMC, UP HD camera, and AI Core module. Besides the hardware, it comes with a complete software package including Ubuntu 16.04.4 Desktop, OpenVINO™ toolkit, Open CL/ Movidius Driver, Arduino Create and Intel System Studio.

Leveraging a wide range of platforms based on Intel technology alongside extensive AI partners, UP offers solutions to power artificial intelligence and edge computing through Hailo-8™ M.2 AI Acceleration Module together with UP platforms.

The latest Hailo-8 M.2 AI module delivers unprecedented AI performance for UP edge devices with high power efficiency Hailo-8™ AI Processor. Power your AI solutions to meet next-generation demands and step up your UP Squared Pro/ UP Squared 6000/ UP Xtreme i11 performance with best-in-class AI processor in a module. Execute in real-time and with low-power deep neural network inferencing for a broad range of market segments.