At the end of a three-hour keynote address for Amazon’s annual re:Invent conference, which is taking place virtually this year, Amazon Web Services chief executive Andy Jassy wrapped up with an extended discussion about edge computing and its role in hybrid computing.
“Hybrid is not just about whether its on-premise or in the cloud,” said Jassy. Instead, IT needs “the same APIs, the same control plane, the same tools, the same hardware they get in AWS regions,” said Jassy. He was referring to Amazon’s AWS Outposts, a rack of equipment deployed at a customer facility that is a fully-managed service from Amazon.
Jassy said Amazon has made the Outposts offering easier to purchase now with new form factors, 1U and 2U rack units, versus an entire rack-size deployment.
In addition to Outposts, Jassy discussed AWS Wavelength, the mobile infrastructure service that Amazon is rolling out in various U.S. cities with Verizon Communications, as part of the 5G infrastructure roll-out. The service had been unveiled last year, and is currently available in eight U.S. cities. Later this year, Amazon will be collaborating with Japanese telco KDDI in Tokyo and with South Korea’s SK Telecom. Next year, the Wavelength infrastructure will be coming to Vodafone’s market in London.
Also: AWS introduces Proton container management service
In addition to the edge computing news, Jassy discussed a plethora of offerings for all areas of infrastructure and applications. In contact centers, Jassy outlined enhancements to the company’s AWS Connect software, called Connect Wisdom. The service has built-in connectors to knowledge repositories and listens into calls and delivers information on customers so agents can personalize their service.
AWS has added industrial machine learning offerings that Jassy discussed, including an appliance computer for on-premise machine vision applications called AWS Panorama Appliance, and Monitron, a sensor gateway appliance to keep tabs on equipment.
In addition, the company unveiled new in-house silicon for developing neural networks, called AWS Trainium. The chip follows other custom AI silicon from AWS, Graviton and Inferentia.
Also: AWS launches Amazon Connect real-time analytics, customer profiles, machine learning tools
The company extended its application container offerings with the Elastic Container Service Anywhere, seeing customers use containers with the AWS control plane in their own data centers. A new control plane, called Proton, allows platform teams to create a “stack” that defines everything needed to provision, deploy, and monitor a service, including compute, networking, code pipeline, security, and monitoring.