Amazon Timestream is a purpose-built time series database service for collecting, storing, and processing time-series data such as server and network logs, sensor data, and industrial telemetry data for IoT and operational applications. Amazon Timestream processes trillions of events per day at one-tenth the cost of relational databases, with up to one thousand times faster query performance than a general-purpose database.

Amazon SageMaker now enables developers and data scientists to quickly and easily develop reinforcement learning models at scale with Amazon SageMaker RL.

The AWS Deep Learning AMIs for Ubuntu and Amazon Linux now support distributed training of TensorFlow deep learning models with near-linear scaling efficiency up to 256 GPUs. The AWS Deep Learning AMIs come pre-built with an enhanced version of TensorFlow that is integrated with an optimized version of the Horovod distributed training framework to provide this level of scalability. With this enhancement, you can now train the ResNet50 model with TensorFlow-Horovod in just under 15 minutes.

Amazon RDS on VMware is a service, now in preview, that delivers Amazon RDS managed relational databases in VMware vSphere on-premises data centers. RDS on VMware automates database provisioning, operating system and database patching, backup, point-in-time restore, storage and compute scaling, instance health monitoring, and failover.

Starting today, you can now add tagging to your Lightsail resources, including virtual servers, managed databases, load balancers, block storage, snapshots, and DNS zones. Lightsail tags allow you to easily organize your projects, create cost allocation reports for billing, and enable access control for your resources.

Developers, start your engines. We are excited to introduce the preview of AWS DeepRacer, the fastest way to get rolling with machine learning, literally. Get hands-on with a fully autonomous 1/18th scale race car driven by reinforcement learning, 3D racing simulator, and global racing league. You can pre-order your AWS DeepRacer now on amazon.com.

AWS IoT Greengrass now supports Amazon SageMaker Neo. Neo enables machine learning models to train once and run anywhere in the cloud and at the edge. Neo automatically optimizes TensorFlow, MXNet, PyTorch, ONNX, and XGBoost models for deployment on ARM, Intel, and Nvidia processors. Optimized models run up to twice as fast and consume less than a tenth of the memory footprint. Neo will also be available as open source code under the Apache Software License soon, enabling hardware vendors to customize it for their processors and devices. Using Neo with AWS IoT Greengrass, you can retrain these models in Amazon SageMaker, and update the optimized models quickly to improve intelligence on these edge devices. You can use a broad range of devices based on the Nvidia Jetson TX2, Arm v7 (Raspberry Pi), or Intel Atom platforms.  

Starting today, you can easily export Lightsail instances and volumes to EC2 with a simple, guided experience. With this feature, Lightsail offers an additional way to grow your applications and scale your cloud deployments by utilizing the full selection and configurability of EC2.

AWS App Mesh is a service mesh that allows you to easily monitor and control communications across microservices applications.

You can now hibernate your Amazon EC2 instances backed by Amazon EBS and resume them at a later time. Applications can pick up exactly where they left off instead of rebuilding the memory footprint all over again. For example, applications that rely on caches and other memory-centric components can take tens of minutes to preload or warm up. These factors impose a delay and can force you to over-provision in case you need incremental capacity very quickly. Using hibernate, you can maintain a fleet of pre-warmed instances with memory footprint that can get to a productive state faster. You can do this without modifying your existing applications. Hibernate is just like closing and opening your laptop lid, with your application starting up right where it left off.

Upon hibernation, your instance’s EBS root volume and any other attached EBS data volumes are persisted between sessions. Additionally, data from memory (RAM) is also saved to your EBS root volume. Upon resume, your EBS root device is restored from its prior state, including the RAM content. Previously attached data volumes are reattached and the instance retains its instance ID. While the instances are in hibernation, you pay only for the EBS volumes and Elastic IP addresses attached to it.

This feature is available for On-Demand and Reserved Instances running on freshly launched M3, M4, M5, C3, C4, C5, R3, R4, and R5 instances running Amazon Linux 1. The AMI snapshot used to launch the instance must be encrypted. This ensures protection of sensitive contents in memory (RAM) as they get copied to the root volume.

You can enable hibernation for your EBS-backed instances at launch. You can then hibernate and resume your EBS-backed EC2 instances through the AWS Management Console , or though the AWS SDK and CLI using the existing stop-instances and start-instances commands.

EC2 Instance Hibernation is now available in the US East (N. Virginia, Ohio), US West (N. California, Oregon), Canada (Central), South America (Sao Paulo), Asia Pacific (Mumbai, Seoul, Singapore, Sydney, Tokyo), and EU (Frankfurt, London, Ireland, Paris) Regions.

To learn more about hibernation, visit this blog . For information about enabling hibernation for your EC2 instances, visit our FAQs or technical documentation .