Chip companies partner with Foxconn for edge AI in video applications

Chip companies partner with Foxconn for edge AI in video applications

A trio of edge computer vendors including Foxconn Technology Group have partnered to create a new product for efficient AI processing for video analysis in edge deployments.

AI chipmaker Hailo Technologies Ltd. and Socionext Inc., which makes systems-on-a-chip hardware for use in video and imaging products, are incorporating their technology into Foxconn’s BOXiedge server platform.

The incorporation of AI-specific functions on a specialized processor provides significant energy efficiency for standalone artificial intelligence inference nodes. A media release from Socionext indicates that the Boxiedge is primarily aimed at smart city, medical, retail and industrial Internet of Things applications.

The BOXiedge is a high-density and fanless edge device from Taiwan industrial giant Foxconn. Socionext, based in Japan, provides SoC (system-on-chip) technology and brings its parallel multi-core processor SynQuacer SC2A11 to the partnership. Israel-based Hailo contributed its Hailo-8 deep learning processor, which boasts the ability to process up to 26 tera operations per second (TOPS).

Socionext

Buyers can shunt at least 20 camera streams through the video management system (VMS_ server for processing and analysis on the edge, according to Socionext. Capabilities also include pose estimation and image classification.

Foxconn itself has used AI solutions developed in-house that use video feeds to detect product defects on different production lines, leading to an improvement in reporting accuracy from 95% to 99% and a reduction of at least one-third of the operating costs for appearance defect inspection projects. Customers of the newly combined solution now have a way to implement these kinds of projects without having to do AI model development and training themselves.

The partnership comes as a flood of products and investment comes for AI-focused chip designs. Hailo recently closed $60m in funding to compete in the edge AI hardware market, a landscape already dotted with aggressive and well-funded chip startups including Graphcore, Xnor, AIStorm, Esperanto Technologies, Quadric and Flex Logic.

Established companies like Sony aren’t being left out. Sony recently announced a chip design that combines AI processing capabilities directly onto the video sensor, for example.

Article Topics

 |   |   |   |   |   | 

Comments

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Sponsored Links

Avassa: Empowers companies to bridge the gap between modern containerized applications development and operations and distributed edge infrastructure. https://avassa.io/

DataBank: We believe there is a different edge to be served - the “middle edge" - that will become the first step for many in their journey to the edge. https://www.databank.com/

Latitude.sh: Where the power of bare metal meets the flexibility of the cloud. Deploy physical servers across 23 global locations in as little as 5 seconds. https://www.latitude.sh/

Zenlayer: A massively distributed edge cloud service provider operating over 270 PoPs around the world, with expertise in fast-growing emerging markets. https://www.zenlayer.com/

OnLogic: A global industrial PC manufacturer and solution provider focused on hardware for IoT and edge AI, OnLogic designs highly-configurable computers engineered for reliability. https://www.onlogic.com/

Featured Company

Latest News