Lattice Blog


Latest Consumer News from Lattice

Consumer News from Lattice
Posted 11/01/2019 by Lattice Semiconductor

Posted in

Enhanced sensAI Stack Enables More Accurate, Lower Power Edge Solutions

Designers trying to pack more performance into their AI edge apps have a new tool at their disposal. Say hello to the latest rev of Lattice’s sensAI solutions stack, the award-winning tool OEMs use to develop Artificial Intelligence (AI) and Machine Learning (ML) experiences for next-gen smart devices targeted at industrial, automotive, compute and consumer applications. The new stack offers a wide range of performance improvements for designers using Lattice’s iCE40 UltraPlus and ECP5 FPGAs including improved reference designs, support for more compact and efficient neural network models, and deeper quantization to accommodate larger models for processing higher resolution and/or faster frame-per-second images in vision applications. The new reference designs allow designers to more easily create popular AI experiences such as key phrase detection and human face recognition. And running the sensAI stack on the iCE40 UltraPlus and ECP5 FPGAs allows designers to more easily add AI applications to new and existing product designs.

Free Webinar Explores Low Power AI Design

sensAI Webinar

Ever wonder how difficult it would be to develop AI solutions that operate on just a few milliwatts (mW)? Now you can find out. Just check out the new webinar presented by Lattice Semiconductor. Titled “Delivering milliWatt AI to the Edge with Ultra Low Power FPGAs,” the first part of the presentation explores Lattice’s sensAI stack, various system-level architectures to implement smart edge devices, and Lattice’s end-to-end reference design flow, while the second portion focuses on memory management techniques, quantization and fractional settings used for deployment of small models on low power edge devices. To meet the needs of designers worldwide, the one-hour webinar will be presented at 4 different times, in English, Chinese, and Japanese.

  • November 19th, 2019 | 11:00 AM – 12:00 PM PST | English
  • November 21st, 2019 | 3:00 PM – 4:00 PM CET | English
  • November 26th, 2019 | 2:00 PM – 3:00 PM JST | Japanese
  • November 26th, 2019 | 2:00 PM – 3:00 PM CST |Chinese

Blog Entries Cover Growing Range of Topics

From embedded vision to Artificial Intelligence (AI), recent posts to Lattice’s company blog offer useful insight into a wide range of hot topics. “CrossLinkPlus: Embedded Vision at Flash Speed,” for example, examines three trends OEMs need to address as they build embedded vision systems. “sensAI Empowers AI Innovation at Microsoft Hackathon 2019” describes how a Lattice-sponsored section of the Hackathon 2019 explored the use of sensors and AI. A third blog entry, “Dealing with Rising Performance Demands at the Edge” takes a look at how the latest version of Lattice’s sensAI technology stack helps designers meet rising performance needs. Finally, a fourth blog entry, “Watch the Lattice sensAI Solutions Stack Deliver Low Power Smart Vision to the Edge”, shows the reader how recent improvements to the sensAI stack are making low power smart vision a reality on the network edge.

CrossLink FPGAs Simplify Use of Embedded Vision

CrossLink Webinar

Want to learn how to add support for the Mobile Industry Processor Interface (MIPI) in Embedded Vision in applications like smart cars, industrial robotics and drones? Check out Lattice’s latest webinar entitled “Supporting MIPI in Embedded Vision System Designs Using FPGAs.” This overview tells you all you need to know about the challenges you’ll face. At the same time, it lays out how you can use Lattice’s CrossLink FPGAs, its easy-to-use software and hardware tools, and its pre-verified IP blocks and reference designs to quickly get up to speed. The webinar is available in English, Chinese and Japanese.