OVERVIEW
Four Routes to AI Inference at the Edge
Whether the application calls for raw parallel throughput, extreme power efficiency, a purpose-built AI platform, or an upgrade to hardware already in the field, New Era Electronics configures and supports the right approach — with the long lifecycles and controlled BOMs that OEM production requires.
01 / BUILT FOR EXTREME PARALLEL PROCESSING POWER
GPU Computers
GPU computers leverage their thousands of GPU cores and the CUDA platform from NVIDIA to boost the parallel processing of the device further than a normal CPU ever could. This enables the device to perform millions of simultaneous operations necessary for applications such as video analytics, deep learning, machine vision and more.
02 / PURPOSE BUILT. INDUSTRIAL. POWER EFFICIENT
VPU Computers
Computers which leverage Vision Processing Units, a type of expansion card that is application specific. Intel VPUs are purpose-built to boost deep learning performance in fields such as computer vision, automatic speech recognition and natural language processing. Compared with general purpose GPUs, VPUs are more efficiently designed for AI and as such are lower cost and draw significantly less power — allowing for a smaller footprint and fanless design. Ideal for IoT edge deployments.
03 / A COMPLETE SYSTEM, DESIGNED BOTTOM UP FOR AI
NVIDIA Jetson Computers
NVIDIA Jetson computers are a complete package designed for AI. They do not take an expansion card to add AI functionality — rather, the entire platform was designed from the beginning for AI embedded and edge devices. Jetson systems come pre-designed and configured with a CPU, a GPU with specially designed AI cores, memory, power management and more, with a low power, fanless industrial design.
04 / BOOST THE AI PROCESSING PERFORMANCE OF YOUR DEVICE
AI Accelerators
Intel® Movidius™ Myriad™ X VPUs designed to boost the AI capabilities of compatible computers. Intel VPUs are supported by OpenVINO, the open source toolkit developed by Intel specifically for optimizing and deploying AI inference, and by frameworks such as TensorFlow, Caffe, and MXNet. Available in Mini PCIe, M.2 and full-sized PCIe for all levels of performance.
APPLICATIONS READY
AI Computers Serve a Wide Array of Industries and Applications
ANPR
Automatic number plate recognition is widely used in public transportation, toll roads, road safety and even autonomous vehicles. With more than 2 billion vehicles in use worldwide, it is becoming more and more of a necessity. AI image processing powered by VPU and GPU computers is essential for training models that ensure accuracy despite factors such as lighting, motion blur, weather and viewing angle.
Machine Vision
Machine vision has long been a staple in automation and manufacturing. As technology advances, applications for machine vision have expanded to autonomous vehicles, kiosks and security. The parallel processing afforded by GPU and VPU computers allows for more compute-intensive vision processing to fulfill these needs.
Defect Detection
The field of defect detection has evolved from simple rules-based computation to highly complex detection and prediction based on deep learning algorithms. To make full use of this evolving technology New Era provides high-powered industrial AI computers which can accept GPUs and VPUs.
Video Analytics
Video analysis covers a wide range of applications such as facial recognition, security, demographics, object recognition, and more. New Era provides a wide variety of VPU and GPU computers suitable for deployment in a wide range of environments and sizes.
Smart Retail
Customer count, customer flow, advertisement metrics, smart forecasting, and virtual attendants are just a few of the features of a fully implemented smart retail system. There are a great many ways that brick and mortar stores are leveraging AI computers to improve customer experience and profitability.
Autonomous Vehicle
Whether it's a self-guided vehicle for passenger transport, an autonomous drone, or an autonomous mobile robot for use in a warehouse, autonomous vehicles are a rapidly evolving field that leverages AI to its fullest. New Era carries a line of AI computers which are also rugged and in-vehicle certified.
Deep Learning
Deep learning, much like many other AI applications, is one whose performance is greatly improved by greater graphics and computing power. New Era carries several lines of industrial GPU and VPU computers compatible with a long list of machine learning libraries and frameworks.
IoT Edge
The essence of IoT edge computing is taking load off of centralized servers by performing computing tasks in the field, closer to the point of data input. The benefits of this are faster analysis of data, less load on cloud servers, and real-time prediction. Depending on the nature of the data and analysis, AI and GPU computers can be a perfect fit for this application.
