AWS re:Invent Day 1: Top 5 announcements include machine learning, storage innovations, and container capabilities

CEO Andy Jassy covered 24 new product announcements in his three-hour keynote on the first day of the virtual Amazon Web Services event.

The newest products and services from Amazon Web Services (AWS) extend cloud tools and services to on-prem installations, reduce the cost of machine learning operations, and create more storage options. AWS CEO Andy Jassy announced 24 new capabilities during the keynote at the first day of AWS re:Invent which is a virtual event this year. The news covered everything from customer service platforms and computer vision algorithms to machine learning operations and serverless deployments.

Jassy said the company has been focused on listening to customers and inventing more options for instances, containers, and serverless deployments. Reducing costs is always a priority as well, he said, so AWS engineers looked for ways to make common operations more cost efficient.

Here is a look at all the product news AWS announced today.

More support for machine learning

Jassy said that machine learning has been growing rapidly, and a lot of the costs are related to inference, not necessarily training the data initially. He said that cloud providers have not focused on reducing those costs. 

“Alexa has reduced their cost of inference by 30% and lowered their latency by 25% using our Inf1 instances,” he said.

SEE: Cloud data storage policy (TechRepublic Premium)

AWS is also working on reducing costs for training models with Habana Gaudi-based Amazon EC2 instances that will be available in 2021. AWS Trainium is another option for reducing costs and will be available in the second half of 2021. 

The company also announced five new industrial machine learning services:

• Monitron for end-to-end machine monitoring that includes sensors, gateway, and machine learning service to detect abnormal equipment conditions that may require maintenance
• Lookout for Equipment to give customers with existing equipment sensors the ability to use AWS machine learning models to detect malfunctions
• AWS Panorama Appliance for customers with existing cameras in industrial facilities to use computer vision to improve quality control and workplace safety
• AWS Panorama Software Development Kit (SDK) to allow industrial camera manufacturers to embed computer vision capabilities in new cameras
• Lookout for Vision to use AWS-trained computer vision models on images and video streams to find flaws in products or processes

Axis, ADLINK Technology, BP, Deloitte, Fender, GE Healthcare, and Siemens Mobility are using these new machine learning services, according to a press release.

Bringing cloud container capabilities on prem

Jassy claimed that two thirds of the containers in the cloud run on AWS. AWS has three offerings for containers: Elastic Kubernetes Service, Elastic Container Service, and Fargate.

He said all three offerings continue to grow like weeds, and that many customers use all three services to accommodate a particular team or use case.

Jassy said customers have requested options to manage containers on premises as they make the transition to the cloud. The four new container capabilities announced today extend the same cloud tools to on-prem containers:

• ECS Anywhere enables customers to run Amazon Elastic Compute Services in their own data centers
• Amazon EKS Anywhere provides the ability to run Amazon Elastic Kubernetes Services in their own data centers
• AWS Proton provides a new service to automate container and serverless application development and deployment
• Amazon Elastic Container Registry Public provides developers a way to share and deploy container software publicly

New storage innovations

In his keynote presentation, Jassy emphasized the need to constantly reinvent products and services and to build what customers want. He said Amazon customers have been asking for more options for global content distribution, storage compliance, and data-sharing. These four storage innovations announced Tuesday meet those needs: 

• Amazon EBS io2 Block Express volumes: A storage area network (SAN) built for the cloud, with up to 256,000 IOPS, 4,000 MB/second throughput, and 64 TB of capacity
• Next-generation Amazon EBS Gp3 volumes: A new iteration that gives customers the ability to provision additional IOPS and throughput performance independent of storage capacity and is priced 20% lower per GB than previous generation volumes
• Amazon S3 Intelligent-Tiering: Now includes S3 Glacier Archive and Deep Archive access to existing Frequent and Infrequent access tiers to automatically reduce storage costs for objects that are rarely accessed
• Amazon S3 Replication (multi-destination): The ability to replicate data to multiple S3 buckets simultaneously in the same AWS Region or any number of AWS Regions 

Supporting a move to serverless installations

Aurora is the fastest growing service in the history of AWS, Jassy said in announcing the next version of Amazon Aurora Serverless. 

SEE: Top cloud providers in 2020: AWS, Microsoft Azure, and Google Cloud, hybrid, SaaS players (TechRepublic)

This new capability is designed in part to make it easier to migrate from SQL Server to Amazon Aurora. This new iteration includes these features:

• Aurora Serverless v2 to accommodate hundreds of thousands of transactions in a fraction of a second, delivering up to 90% cost savings compared with provisioning for peak capacity
• Babelfish for Aurora PostgreSQL to provide the ability to run SQL Server applications directly on Amazon Aurora PostgreSQL with little to no code changes 
• Open source Babelfish for PostgreSQL to extend the benefits of the Babelfish for Amazon Aurora PostgreSQL translation layer which will be coming in 2021 under the permissive Apache 2.0 license in GitHub

Expanding analytics capabilities

Rahul Pathak, vice president of analytics at AWS, said in a press release that these new services represent an order-of-magnitude performance improvement for Amazon Redshift. These new services make it easier for customers to move data between data stores and to ask natural language questions in their business dashboards and receive answers in seconds, according to the company. The new options include:

• Advanced Query Accelerator for Amazon Redshift provides a new hardware-accelerated cache to improve query performance  
• AWS Glue Elastic Views lets developers build materialized views that automatically combine and replicate data across multiple data stores
• Amazon QuickSight Q is a machine learning-powered capability lets users type questions about their business data in natural language and receive highly accurate answers in seconds

Five new capabilities for Amazon Connect 

Amazon Connect is the company’s customer service-in-a-box offering. John Hancock, Capital One, Intuit, Best Western, Fujitsu, GE Appliances, and Square use the service to make customer service more efficient, according to AWS. The new features announced today include: 

• Connect Wisdom provides contact center agents with real-time information 
• Connect Customer Profiles gives agents a unified profile of each customer  
• Real-Time Contact Lens offers a new option for contact center managers to use during a call
• Connect Tasks automates, tracks, and manages tasks for contact center agents
• Connect Voice ID delivers real-time caller authentication using machine learning-powered voice analysis

Also see