7/10/2025, 12:00:00 AM ~ 7/11/2025, 12:00:00 AM (UTC)

Recent Announcements

Amazon SageMaker HyperPod accelerates open-weights model deployment

Amazon SageMaker HyperPod now supports deploying both open-weights foundation models from Amazon SageMaker JumpStart and your own fine-tuned models from Amazon S3 and Amazon FSx directly to Amazon SageMaker HyperPod. This enables you to seamlessly train, fine tune, and deploy models on the same HyperPod compute resources, maximizing resource utilization across the entire model lifecycle\n In a few quick steps, you can choose an open-weights foundation model from SageMaker JumpStart and quickly deploy it on your SageMaker HyperPod cluster. SageMaker automatically provisions the infrastructure, deploys the model on your cluster, enables auto-scaling, and configures the SageMaker endpoint. SageMaker scales the compute resources up and down through HyperPod task governance as the traffic on model endpoints changes, and automatically publishes metrics to the HyperPod observability dashboard to provide full visibility into model performance. You can deploy models from SageMaker JumpStart in all AWS Regions where HyperPod is available: US East (N. Virginia), US West (N. California), US West (Oregon), Asia Pacific (Mumbai), Asia Pacific (Singapore), Asia Pacific (Sydney), and Asia Pacific (Tokyo), Europe (Frankfurt), Europe (Ireland), Europe (London), Europe (Stockholm), and South America (São Paulo). To learn more, visit SageMaker HyperPod webpage, blog, and documentation.

Amazon SageMaker Studio now supports remote connections from Visual Studio Code

Today, AWS announces remote connection from Visual Studio Code to Amazon SageMaker Studio development environments, enabling AI developers to use Visual Studio Code with SageMaker AI’s scalable compute resources. This new capability enables developers to connect from Visual Studio Code to SageMaker Studio in minutes instead of hours, enabling them to rapidly scale their model development.\n SageMaker Studio offers a broad set of fully managed cloud interactive development environments (IDE), including JupyterLab and Code Editor based on Code-OSS (VS Code - Open Source). Starting today, you can use your customized local VS Code setup, including AI-assisted development tools and custom extensions, while accessing SageMaker AI’s compute resources and your data. You can authenticate using the AWS Toolkit extension in VS Code or through SageMaker Studio’s web interface. Once authenticated, connect to any of your SageMaker Studio development environments in a few simple clicks. You maintain the same security boundaries as SageMaker Studio’s web-based environments while developing AI models and analyzing data in Visual Studio Code.

This feature is now available in US East (Ohio) Region.

To learn more, see the following resources:

Amazon SageMaker Studio webpage

AWS ML Blog

Documentation

Amazon SageMaker HyperPod introduces CLI and SDK for AI Workflows

We are excited to announce the general availability of the Amazon SageMaker HyperPod Command Line Interface (CLI) and Software Development Kit (SDK). These tools make it easier for developers and ML practitioners to build, train, and deploy large-scale AI models on SageMaker HyperPod.\n SageMaker HyperPod CLI offers a simple, consistent command-line experience for managing HyperPod clusters and quick experimentation. The SDK offers intuitive programmatic access to HyperPod’s distributed training and inference capabilities and enables developers to have granular control over their workload configurations. With these tools, data scientists and ML engineers can easily launch training jobs, deploy scalable inference endpoints, and monitor cluster performance. Using simple commands, customers can access system logs and HyperPod observability dashboards, enabling them to debug issues and accelerate model development. These new developer interfaces enable customers to build and deploy production generative AI models faster on SageMaker HyperPod. The HyperPod CLI and SDK are available in all AWS commercial regions where SageMaker HyperPod is supported. To get started with the SageMaker HyperPod CLI and SDK, visit the SageMaker HyperPod Developer Guide.

AWS announces 100G expansion in Kolkata, India

Today, AWS announced the expansion of 100 Gbps dedicated connections at the AWS Direct Connect location in the STT Kolkata, DC1 data center center near Kolkata, India. You can now establish private, direct network access to all public AWS Regions (except those in China), AWS GovCloud Regions, and AWS Local Zones from this location. This is the third AWS Direct Connect location in India to provide 100 Gbps connections with MACsec encryption capabilities.\n The Direct Connect service enables you to establish a private, physical network connection between AWS and your data center, office, or colocation environment. These private connections can provide a more consistent network experience than those made over the public internet. For more information on the over 142 Direct Connect locations worldwide, visit the locations section of the Direct Connect product detail pages. Or, visit our getting started page to learn more about how to purchase and deploy Direct Connect.

Fully managed MLflow 3.0 now available on Amazon SageMaker AI

Amazon SageMaker now offers fully-managed support for MLflow 3.0 that streamlines AI experimentation and accelerates your GenAI journey from idea to production. This release transforms managed MLflow from experiment tracking to providing end-to-end observability, reducing time-to-market for generative AI development.\n As customers across industries accelerate their generative AI development, they require capabilities to track experiments, observe behavior, and evaluate performance of models and AI applications. Data scientists and developers lack tools for analyzing the performance of models and AI applications from experimentation to production, making it hard to root cause and resolve issues. Teams spend more time integrating tools than improving their models or generative AI applications. With this launch, fully managed MLflow 3.0 on Amazon SageMaker AI enables customers to accelerate generative AI by making it easier to track experiments and monitor performance of models and AI applications using a single tool. Tracing capabilities in fully managed MLflow 3.0 enable customers to record the inputs, outputs, and metadata at every step of a generative AI application, helping developers quickly identify the source of bugs or unexpected behaviors. By maintaining records of each model and application version, fully managed MLflow 3.0 offers traceability to connect AI responses to their source components, allowing developers to quickly trace an issue directly to the specific code, data, or parameters that generated it. This dramatically reduces troubleshooting time and enables teams to focus more on innovation.

Fully managed MLflow 3.0 on Amazon SageMaker AI is now available in all regions where Amazon SageMaker is offered, excluding China Regions and GovCloud (US) Regions.

To learn more about fully managed MLflow 3.0 on Amazon SageMaker AI, visit the Amazon SageMaker developer guide.

Amazon SageMaker HyperPod announces new observability capability

Amazon SageMaker HyperPod’s new observability capability allows customers to accelerate generative AI model development by providing comprehensive visibility across compute resources and model development tasks. It takes away the manual work of collecting hundreds of metrics from across the stack, visualizing the correlations between them, and restoring the generative AI model development task performance. HyperPod observability tracks task performance metrics in real-time, alerts customers when any of them deteriorate, and automatically remediates the root cause with customer-defined policies.\n SageMaker HyperPod observability transforms how customers monitor and optimize their generative AI model development tasks. Through a unified dashboard pre-configured in Amazon Managed Grafana with the monitoring data automatically published to an Amazon Managed Prometheus workspace, customers can now see generative AI task performance metrics, resource utilization, and cluster health in a single view. This allows teams to quickly spot bottlenecks, prevent costly delays, and optimize compute resources. Customers can define automated alerts, derive use-case specific task metrics, and publish them to the unified dashboard with just a few clicks. By reducing troubleshooting time from days to minutes, this capability helps customers accelerate their path to production and maximize the return on their AI investments.

SageMaker HyperPod observability is available in all AWS Regions where SageMaker HyperPod is supported, except US West (N. California) and Asia Pacific (Melbourne). To learn more and get started, visit the blog, documentation, and SageMaker HyperPod webpage.

Anthropic’s Claude 3.7 Sonnet is now available on Amazon Bedrock in AWS GovCloud (US-West)

Anthropic’s Claude 3.7 Sonnet hybrid reasoning model is now available in AWS GovCloud (US-West). Claude 3.7 Sonnet offers advanced AI capabilities with both quick responses and extended, step-by-step thinking made visible to the user. This model has strong capabilities in coding and brings enhanced performance across various tasks, like instruction following, math, and physics. Anthropic’s Claude 3.7 Sonnet model is also FedRAMP High and Department of Defense Cloud Computing Security Requirements Guide (DoD CC SRG) Impact Level (IL) 4 and 5 approved within Amazon Bedrock in the AWS GovCloud (US) Regions.\n Claude 3.7 Sonnet introduces a unique approach to AI reasoning by integrating it seamlessly with other capabilities. Unlike traditional models that separate quick responses from those requiring deeper thought, Claude 3.7 Sonnet allows users to toggle between standard and extended thinking modes. In standard mode, it functions as an upgraded version of Claude 3.5 Sonnet. In extended thinking mode, it employs self-reflection to achieve improved results across a wide range of tasks. Amazon Bedrock customers can adjust how long the model thinks, offering a flexible trade-off between speed and answer quality. Additionally, users can control the reasoning budget by specifying a token limit, enabling more precise cost management. Claude 3.7 Sonnet is also available on Amazon Bedrock in the Europe (London), Europe (Frankfurt), Europe (Ireland), Europe (Paris), Europe (Stockholm), US East (N. Virginia), US East (Ohio), and US West (Oregon) regions. To get started, visit the Amazon Bedrock console. Integrate it into your applications using the Amazon Bedrock API or SDK. For more information, see the AWS News Blog, Claude in Amazon Bedrock and the Amazon Bedrock documentation.

AWS Blogs

AWS Japan Blog (Japanese)

AWS Open Source Blog

AWS Cloud Financial Management

AWS Big Data Blog

AWS Database Blog

AWS HPC Blog

Artificial Intelligence

AWS for M&E Blog

AWS Security Blog

Open Source Project

Amazon EKS Anywhere