Lattice claims low power FPGA will enable mass market AI in edge devices
Lattice Semiconductor believes its sensAI technology stack, combining modular hardware kits, neural network IP cores, software tools, reference designs and custom design services will open the door for rapid deployment of machine learning inference across IoT applications that require low power consumption.
The sensAI is optimised to provide low power operation (one mW to one W), small size, and production-priced (~one to $10) benefits of ASICs, with the flexibility of FPGAs, to support evolving algorithms, interfaces, and tailored performance, says Lattice.
These attributes are expected to accelerate deployment of AI into fast growth consumer and industrial IoT applications including mobile, smart home, smart city, smart factory, and smart car products.
The package size is 5.5 to 100mm2, while interface flexibility, with MIPI CSI-2, LVDS and GigE options, and the high-volume pricing offer a fast-track implementation of edge computing close to the source of data, says Lattice.
Deepak Boppana, senior director, product and segment marketing at Lattice Semiconductor said: “By delivering a full-featured machine learning inferencing technology stack combining flexible, ultra-low power FPGA hardware and software solutions, the Lattice sensAI stack accelerates integration of on-device sensor data processing and analytics in edge devices.” He goes on to say that the company’s expertise in FPGA technology are the basis for new edge computing solutions for edge connectivity, implementing flexible sensor interface bridging and data aggregation in high-volume IoT applications”, he continued, citing smart speakers, surveillance cameras, industrial robots and drones as potential applications.
“The edge is getting smarter with more computing capabilities being deployed for real-time processing of data from an expanding range of sensors, as seen in the consumer IoT space, and the emergence of artificial intelligence is only accelerating this trend,” said Michael Palma, research director at IDC.
With the adoption of machine learning technology, latency, privacy and network bandwidth limitations are pushing computing to the edge. IHS Markit expects 40 billion IoT devices at the edge between 2018 and 2025, and predicts that in the next five to 10 years, the convergence of IoT, AI-based edge computing and cloud analytics will disrupt “each and every industry vertical and domain, as well as to foster new business opportunities”. Semico Research predicts unit growth for edge devices with AI will increase by more than 110 per cent CAGR over the next five years.
Lattice’s sensAI stack includes modular hardware platforms, the ECP5 device-based video interface platform (VIP), including the Embedded Vision development kit, and iCE40 UltraPlus device-based Mobile Development Platform (MDP).
The IP core include the Convolutional Neural Network (CNN) accelerator and Binarised Neural Network (BNN) accelerator. Software tools are the neural network compiler tool for Caffe/TensorFlow to FPGA, Lattice Radiant design software and Lattice Diamond design software.
Face detection, key phrase detection, object counting, face tracking, and speed sign detection reference designs are also included, with design services from partners for market applications covering smart cities, smart homes and smart factories.