Amazon SageMaker Model Monitor is a new capability of Amazon SageMaker that continuously monitors machine learning (ML) models in production, detects deviations such as data drift that can degrade model performance over time, and alerts you to take remedial actions.

Amazon SageMaker Debugger is a new capability of Amazon SageMaker that provides complete insights into the training process of machine learning (ML) models by automating the capture and analysis of data from training runs at real time, with no code changes.

The AWS Deep Learning Containers are available today with the latest framework versions of Tensorflow 1.15, PyTorch 1.3.1, and MXNet 1.6.0-rc0. You can launch the new versions of Deep Learning Container on Amazon Sagemaker, Amazon Elastic Kubernetes Service (Amazon EKS), self-managed Kubernetes on Amazon EC2, and Amazon Elastic Container Service (Amazon ECS). For a complete list of frameworks and versions supported by the AWS Deep Learning Containers, see release notes.

Amazon Elastic Container Service (ECS) Capacity Providers are now available. Capacity Providers are a new way to manage compute capacity for containers, that allow the application to define its requirements for how it uses the capacity. With Capacity Providers, you can define flexible rules for how containerized workloads run on different types of compute capacity, and manage the scaling of the capacity. Capacity Providers improve the availability, scalability, and cost of running tasks and services on ECS.

We are announcing DJL, an open source library to develop Deep Learning models in Java. DJL offers user-friendly APIs to train, test, and deploy Deep Learning models. If you are a Java user interested in Deep Learning, DJL is a great way to start your journey. If you’re a Java developer working with Deep learning models, DJL will simplify the way you train and run predictions.

Amazon Elastic Container Service (ECS) Cluster Auto Scaling is now available. With ECS Cluster Auto Scaling, your ECS clusters running on EC2 can automatically scale as needed to meet the resource demands of all tasks and services in your cluster, including scaling to and from zero. Managed scaling with ECS Cluster Auto Scaling improves the reliability, scalability, and cost of running containerized workloads on ECS. 

Amazon ECS, Amazon EKS, and AWS App Mesh now support AWS Outposts, a fully managed service that extends AWS infrastructure, and tools to virtually any datacenter, co-location space, or on-premises facility for a consistent hybrid experience.  

Amazon Chime now hosts meetings in 14 AWS Regions and 32 Availability Zones. Amazon Chime now automatically selects the region that provides a low latency for all meeting participants, whether they are all local in one place or distributed globally. Meeting region selection is performed for every meeting based on participant proximity to AWS Regions and AWS network telemetry.

The AWS Nitro System is the underlying platform for the latest generation of EC2 instances that enables AWS to innovate faster, further reduce cost for our customers, and deliver added benefits like increased security and new instance types. With the latest set of enhancements to the Nitro system, all new C5/C5d/C5n, M5/M5d/M5n/M5dn, R5/R5d/R5n/R5dn, and P3dn instances now support 36% higher EBS-optimized instance bandwidth, up to 19 Gbps. Also, 6, 9, and 12 TB Amazon EC2 High Memory instances can now support 19 Gbps of EBS-optimized instance bandwidth, a 36% increase from 14 Gbps. 

Today, we are announcing EBS direct APIs for Snapshots that provide read access to EBS snapshots. Backup providers can now easily track incremental changes on EBS volumes via EBS snapshots and streamline their workflows to reduce backup times by up to 70%. This enables them to provide more granular recovery point objectives (RPOs) to AWS customers at lower costs.