Amazon Redshift, the most widely used cloud data warehouse, announces general availability of Automatic Table Optimization, a new self-tuning capability that optimizes the physical design of tables by automatically setting sort and distribution keys to improve query speed. You can use Automatic Table Optimization to get started with Amazon Redshift easily or optimize production workloads by decreasing the administrative effort required to get the best possible performance. 

Amazon Redshift, a fully-managed cloud data warehouse, now supports native integration with select AWS partners from within the Amazon Redshift Console. With the new console partner integration, you can accelerate data onboarding and create valuable business insights in minutes by integrating with select partner solutions. With these solutions, you can bring data from applications like SalesForce, Google Analytics, Facebook Ads, Slack, Jira, Splunk, and Marketo into your Amazon Redshift data warehouse in an efficient and streamlined way. It also enables you to join these disparate datasets and analyze them together to produce actionable insights.

We are excited to announce the availability of Amazon EC2 G4ad instances that provide the best price performance for graphics intensive applications in the cloud. G4ad instances are powered by AMD Radeon Pro V520 GPUs and second-generation AMD EPYC processors, and provide up to 45% better price performance over G4dn instances for graphics intensive applications such as virtual workstations, game streaming, and graphics rendering. 

Today we are announcing the public preview of EMR Studio, an integrated development environment (IDE) that makes it easy for data scientists and data engineers to develop, visualize, and debug data engineering and data science applications written in R, Python, Scala, and PySpark. EMR Studio provides fully managed Jupyter Notebooks, and tools like Spark UI and YARN Timeline Service to simplify debugging.

AWS Global Accelerator announces custom routing accelerator, a new type of accelerator that allows you to use your own application logic to route user traffic to a specific Amazon EC2 destination, while still leveraging the benefits of Global Accelerator. 

Amazon SageMaker Edge Manager is a new feature of Amazon SageMaker that enables developers with the tools for model management across your fleets of edge devices so you can optimize, secure, monitor, and maintain machine learning models on fleets of edge devices such as smart cameras, robots, personal computers, and mobile devices.  

We’re excited to announce new capabilities with Amazon SageMaker Debugger with real-time monitoring of system resources for efficient utilization. With these new capabilities, you can now get automatic recommendations to re-allocate resources for your training jobs, helping you train better and reduce time and costs.

Amazon Lookout for Metrics uses machine learning (ML) to detect anomalies, or unexpected changes in your metrics, helping you proactively monitor the health of your business, diagnose issues and find opportunities quickly – with no ML experience required. 

Today, Amazon announced Amazon Neptune ML, a new capability of Amazon Neptune that uses Graph Neural Networks (GNNs), a machine learning (ML) technique purpose-built for graphs, to make easy, fast, and accurate predictions using graph data. With GNNs, you can improve the accuracy of most predictions for graphs by over 50% when compared to making predictions using non-graph methods based on published research from Stanford University. 

Today we are introducing two new distributed training libraries for Amazon SageMaker, providing integrated methods for you to quickly train large deep learning models. Using partitioning algorithms, these SageMaker distributed training libraries automatically split large deep learning models and training datasets across AWS GPU instances in a fraction of the time it takes to do manually. SageMaker achieves these efficiencies through two techniques: model parallelism and data parallelism. Model parallelism splits models too large to fit on a single GPU into smaller parts before distributing across multiple GPUs to train, and data parallelism splits large datasets to train concurrently in order to improve training speed.