Rapid AI SDK Development Toolkit

Rapid AI SDK Development Toolkit for Quick Wins

100% compatible with select Advantech edges
Cross-platform support for inference

Download Now 〉 Contact Us >

Supercharging AI Inference: Unlocking a 40% Time Savings

Creating an AI-ready development environment can often feel like a formidable challenge. However, Advantech Edge AI SDK, an exceptional toolkit designed to seamlessly support leading AI platforms and their associated inference SDKs, doesn't just simplify the process—it sparks an AI revolution. This SDK empowers developers to effortlessly harness the boundless potential of AI inference.
Unlock Your AI Potential with a Unified Toolset for Any Inference Platform

Imagine one toolkit that simplifies AI inference on any platform. It streamlines your work, saves time, and adds significant value to your AI projects.
The Secret of Delivering AI Inference to Production

The Edge AI SDK empowers top AI platforms and their SDKs, streamlining processes and fueling an AI revolution.
Our Offerings
Inference Benchmark Tool

Inference Runtime SDK

Inference Deployment Platform

Expert Guidance & Support
Inference Benchmark Tool

Developed by ADVANTECH, the whole idea is to present an instant AI experience out-of-the-box to run mainstream face recognition, object detection, and other vison AI models. The user will achieve a good understanding of AI model inference performance before any costly data operations and training take place. After installation, it’s easy to launch AI inference demos with a few clicks. No coding effort or skill is required. Furthermore, there is minimal demand on resources during the evaluation stage. Get started 〉

Benchmark edge AI performance in 3 mins or less

Inference Runtime SDK

To improve the development efficiency of edge AI applications, Advantech provides validated inference SDKs for diverse edge AI hardware. This includes Intel® OpenVINO™, Nvidia TensorRT, and more. More information about other Advantech AI hardware and related AI SDKs will be available here in the near future. Just choose a target device, obtain the compatible AI SDKs, then start developing your AI applications right away. Learn more 〉

Ubuntu LTS OS 〉 Select Advantech AI Edges 〉

Install inference runtime SDKs for validated edge AI platforms

Inference Deployment Platform

Advantech understands what a headache it is to find hardware devices with optimal price and performance when consuming large amounts of energy for AI/ML algorithms and model development. Even if all datasets, algorithms, trainings, UI/UX, and more are functioning, how can you easily deploy an AI application to hundreds, or thousands, of inference devices in a production facility? How can you efficiently manage AI models (software updates, CI/CD), in addition to all remote hardware devices, such as sensors? Learn more 〉

Save 80% of operation time on edge AI deployment & monitoring
Expert Guidance & Support

With Advantech edge AI systems, Advantech can help customers jump into the development stage without wasting time building up their AI environment. It aims to make sure there are no incompatibility issues among drivers, OS, and SDKs. Furthermore, with the close cooperation of Microsoft, Canonical, Nvidia, Intel, and others, these partners will be able to provide further support for technical problems covering embedded operating systems, Azure cloud integration, and even AI application development. Please contact us for more details. Learn more 〉

Global Service Team

Dedicated regional service team to provide local service and support with flexibility and efficiency

System Certification

Ubuntu certification

Nvidia-certified system qualifications

Inference SDKs with high compatibility and troubleshooting assistance

Inference Enablement

Model conversion for inference optimization

Inference platform migration

Benchmarking for AI models vs inference performance

Edge-to-Cloud Collaboration

System monitoring & control by Advantech SUSI

Edge AI DevOps architecture

Azure Cognitive Services for vision, speech AI, and more
Select Edges
- Intel
  EAI-3100
  
  Intel Arc A370M with 8 Xe-cores
  
  PCIe x16 GPU Card Design easy to fit diverse AI applications
  
  Intel® Deep Link and OpenVINO support
  
  Explore >
  ARK-3534
  
  Support Intel® 12th & 13th Gen. Core™ i3/i5/i7/i9 processor
  
  Support up to 4GbE, 8USB, 8COM, 16bit DIO, 2CANBus, TPM2.0
  
  SUSI API and DeviceOn bundled
  
  Explore >
- NVIDIA
  AIR-030
  
  NVIDIA® Jetson AGX Orin 64GB and 32GB Modules
  
  Compact and high performance AI box up to 275 TOPS AI computing
  
  Support Linux Ubuntu 20.04 LTS and JetPack SDK 5.1 and above
  
  Explore >
- HAILO
  AIR-150
  
  Intel 13th gen. Core i3/i5 up to 14 Watt with 10 Cores
  
  Empowered by Hailo-8 M.2 AI module up to 26 TOPS
  
  DeviceOn for device management and remote control supported
  
  Explore >
Vision AI Tutorial
- Vision AI for person and object detection
  
  Scenario:
  
  For production management in a factory, managers analyzed key factors of production monitoring. One is to make sure there is always a WIP for PCB on a production line to be manufactured. The other one is to know if an operator at the target workstation can operate the station correctly and produce normally.
  
  Challenges:
  
  Lack of infrastructure and skills to train the AI model at the office
  
  Limited knowledge to configure the edge computer
  
  Need to document inference results for further analysis
  
  Try it now 〉
- Vision AI for fruit processing and quality classification
  
  Scenario:
  
  To control quality, a fruit distribution packing line manager developed an AI application to detect rotten or stale fruit. He intended to run the AI on Advantech AIR-020 (NVIDIA Jetson Nano) but faced challenges as outlined below.
  
  Challenges:
  
  Lack of infrastructure and skills to train the AI model at the office
  
  Constant updates of AI models at the edge
  
  Real-time monitoring of inference engine performance and results
  
  Try it now 〉
Let’s Run Inferencing Directly on Advantech AI Edge

A benchmark tool with an easy-to-use interface for quick evaluation, along with a verified runtime SDK for efficient development

Get Started 〉
Edge AI SDK FAQ
- What is it?
  
  Edge AI SDK is a toolset that helps customers evaluate, develop and deploy diverse edge AI applications.
- Who is it for?
  
  Edge AI SDK is ideal for AI application developers. Not only software toolkits but also validated edge AI devices can be found here to accelerate the early development stage.
- What can I do with it?
  You can streamline every step of the AI workflow from evaluation, inference development to deployment with Edge AI SDK.
  
  Effortlessly evaluate computing AI performance by several AI models with Inference Benchmark Tool
  
  Efficiently implement AI on Advantech AI systems with Inference Runtime SDK
  
  Remotely manage AI updates on edge systems at scale with Inference Deployment Platform | DeviceOn
- Who provides support?
  
  Edge AI SDK includes enterprise support from Advantech, as the tier-1 service, and AI ecosystem partners, like Nvidia, Intel, Hailo, Microsoft Azure and Canonical, as the tier-2 service.
- How much does it cost?
  
  Only the deployment platform - DeviceOn and advanced consulting services will be charged. More detailed information can be found in corresponding product website or with direct contact.
Ecosystem Partners
Resources