show image

Microsoft’s Satya Nadella puts edge computing at the centre of Build 2018

Microsoft CEO Satya Nadella kicked off the company’s annual Build summit yesterday with a declaration: “The era of the intelligent cloud and intelligent edge is upon us.”

Cloud computing is now an integral part of the software giant’s business. But, like many in the industry, Nadella believes that as smart devices proliferate, the computing process will increasingly take place at the edge of networks – where devices are located – rather than in public or private data centres.

This prediction underscored much of the news on the first day of Microsoft’s annual developer conference. Here’s a roundup of the key “intelligent edge” announcements.

Azure IoT Edge Runtime

Last year, the US tech giant unveiled Azure IoT Edge, a tool allowing edge devices to carry out computing processes locally. Now, it’s open-sourcing Azure IoT Edge Runtime to enable developers to “modify, debug and have more transparency and control” over such applications.

Project Brainwave

Microsoft’s engineers have devised an architecture for neural net processing. The tool, which can be previewed on Azure and the edge, makes Azure “the fastest cloud to run real-time AI”, Microsoft claims.

Custom Vision

Microsoft is developing its Azure Cognitive Services package so that it will run across edge devices over the next few months. The first service is Custom Vision, which will enables drones and industrial equipment to “take critical action quickly without requiring cloud connectivity”, according to the firm.

A partnership with a drone manufacturer

Microsoft has joined forces with the world’s largest drone company, DJI, to create new applications for the unmanned aircraft in agriculture, construction and public safety. Azure IoT Edge and Microsoft’s AI services will support the partnership.

Microsoft teams up with Qualcomm

The firm also announced a partnership with the chipmaker Qualcomm, to build a vision AI developer kit running Azure IoT Edge. The solution has been designed to give developers the hardware and software needed to build camera-based IoT solutions.