Edge AI

Edge AI: Training and inference on the industrial edge

The ability to bring training and inference right to the edge is creating operational benefits in a wide variety of industries, reducing costs, improving data security and increasing performance in machine learning ML models.

Cloud based AI has the horsepower to process vast amounts of data, but when advances in GPU technology are combined with rugged computing hardware, Industrial AIOT applications can be realised right at the edge. Impulse offer a range of rugged edge computers and services to help customers implement their Edge AI projects.

Impulse offer rugged Edge AI compute platforms suitable for installation in wide ranging environments from in-vehicle, industrial, medical, surveilance and drones. We back our hardware solutions with a service layer which includes software integration, hardware customisation and global logistics and project support.

The benefits of Edge AI for industry

Efficiency

Delays in making mission-critical decisions can be costly. With AI inference and GPU computing at the Edge, you can reduce the delay between input and output saving time and money.

Redundancy

Industrial AI placed at the Edge will ensure that decision making and processing continues should connections to systems in the cloud be disrupted by planned or unplanned downtime.

Cost-effectiveness

Minimise costly overheads associated with cloud based services by processing the data at the Edge using GPU computing. Impulse can provide systems to aggregate the data at the source reducing the amount of bandwidth usage and memory allocation in the cloud.

Security

By keeping sensitive data in the local eco-system you can limit the security risks of the public cloud. AI-enabled systems can also detect and deter security intrusions faster and more proactively than humans.

Choosing your rugged Edge AI hardware

NVidia Jetson

Intel Movidius

GPU Cards

NVIDIA Jetson

The NVIDIA Jetson family, comprising the Jetson NANO, TX2, Xavier NX and AGX Xavier , is a series of System-on-Modules (SOM) designed to satisfy a range of applications, from basic Edge AI to fully autonomous machines. To begin your Jetson journey developer kits are available with which you can develop your software. Once ready for deployment on to an edge device, there is a fork in the road. For Industrial applications, using the Jetson System-on-Modules requires carrier board and hardware design capabilities to commercialise an AI solutiuon. Often the favourable option is to use an existing ruggesdised off the shelf solution as this reduces costs and time to market . These pre-packaged Jetson boxes are in rugged cases, designed to meet industrial standards and I/O requirements and are ready to deploy in the field, just add code.

Learn more

Movidius Myriad X VPU

The Movidius Myriad X VPU delivers over 4 trillion operations per second (TOPS) in an ultra low power consumption architecture. An Intel based solution designed and optimised for computer vision applications. Myraid X is available in a range of pluggable cards including M.2, mPCIe and PCIe which makes it an ideal choice to deploy deep learning inference models across intel based Edge Computing products. By utlising the Intel distribution OpenVINO toolkit, pre-trained models in frameworks such as Caffe and Tensorflow can be optimised and output to inference runtimes suitable for the Myriad X VPU. This makes for faster development and deployment to the Edge for applications such as machine vision, surveillance, smart retail and medical imaging.

Discrete GPU Cards

Ruggedised Edge Computers with the ability to take discrete GPU cards are the choice for high performance edge inference applications. These computers allow workstation and datacentre grade PCIe GPU cards to installed into rugged chassis', capable of withstanding shock, vibration and temperature extremes without having to throttle performance. These solutions are ideal for intelligent video analytics, medical imaging, autonomous driving and deep learning machine vision. Impulse build and configure Industrial GPU Computers for wide-ranging customer who need rock solid, reliable and repeatable systems for deployment in harsh environments.

Developed your AI model? Test it on one of our demo systems

We have a number of deployable Aaeon demo units for Edge AI applications available for loan.

Request a demo