IoT Edge Devices
Modern embedded processors enable powerful capabilities for edge applications. These multicore System-on-Chip (SoC) platforms can include ARM processors, GPUs, and other specialized cores – all connected via a high speed memory fabric.
Further, software providers are developing real-time data analytics tools for edge devices. Google TensorFlow can deploy neural networks; Microsoft Azure Stream Analytics on the edge enables streaming SQL processing; and hardware-accelerated OpenCV supports full-rate video analytics.
The convergence of these technology trends is driving the evolution of IoT edge devices towards an edge-to-cloud distributed computing framework. The edge device is part of a continuum in a stream computing topology.
Developing Optimized Embedded Firmware
Developing real-time embedded code for SOCs requires complex toolchains and methods. At Edgespace, we develop tight, optimized code on modern heterogeneous multicore processors, with particular expertise on Intel and NVIDIA platforms. Our embedded technology skills include:
- Linux OS Optimization | Yocto Project
- Embedded Containers (e.g., Docker)
- Custom Kernel Drivers & Sensor Integration
- Real-Time Analytics, including Video Analytics
We have extensive expertise with Microsoft Azure IoT Suite, including custom gateways, IoT Hub integration, device configuration and management, and multiple device-to-cloud communications protocols.
Streamlining the Path to Market
In embedded device design, the line between hardware and software is blurred – the selection of SoC platform and peripheral I/O chips impacts the so-called hardware-software partitioning. At Edgespace, we understand embedded hardware, and can help guide your team through the complex system engineering tradeoffs of device design.
To support your path-to-market, we partner with best-in-class ruggedized device OEMs to provide you manufacturing and lifecycle management solutions.