Kategorie: Amazon AWS

Amazon SageMaker AI now supports serverless model customization for Qwen3.5 models

Amazon SageMaker AI now supports serverless model customization for Qwen3.5, enabling you to fine-tune Qwen3.5 4B, 9B, and 27B parameter models using supervised fine-tuning (SFT) and reinforcement fine-tuning (RFT). Qwen3.5 is a popular open-weight model family from Alibaba Cloud. Before this launch, you could deploy these base models on SageMaker AI and now, you can also adapt them to your specific domains and workflows. 
Model customization enables you to tailor foundation models with your proprietary data so they more accurately reflect your domain knowledge, terminology, and quality standards. Rather than building models from scratch, fine-tuning lets you start from a capable base model and specialize it for your use cases, whether that’s improving accuracy on domain-specific tasks, aligning outputs with your organization’s tone, or improving performance on new tasks using your labeled data. With serverless customization, SageMaker AI handles all infrastructure provisioning and training orchestration, so you can focus on your data and evaluation rather than cluster management, and only pay for what you use.
Serverless model customization for Qwen3.5 on SageMaker AI is available in US East (N. Virginia), US West (Oregon), Asia Pacific (Tokyo), and EU (Ireland). To get started, navigate to the Models page in Amazon SageMaker Studio to launch a customization job, or use the SageMaker Python SDK for programmatic access. To learn more, see the Amazon SageMaker AI model customization documentation.
Quelle: aws.amazon.com

Enhancements to AWS Network Firewall Managed Rules from AWS Marketplace Partners

AWS Network Firewall now supports expanded Managed Rules from AWS Marketplace partners, with new rule group optimizations through partners to include up to 10 million domain name indicators and up to 1 million IP addresses in their managed rule groups. Infoblox is expanding domain name indicators to protect your workloads from critical and high-risk domains. Lumen is introducing new rule groups to stop command and control attacks. ThreatSTOP is adding managed rules for Office of Foreign Assets Control (OFAC) sanctions and expanding global compliance protections with new European Union, Japan, and United Nations sanction coverage. These enhancements give you access to richer, more comprehensive threat intelligence directly within AWS Network Firewall, reducing the operational burden of managing threat feeds and enabling faster, more accurate protection against emerging threats. Whether you need to block malicious domains at scale, defend against command and control infrastructure, or enforce sanctions-based compliance policies, managed rules from AWS Marketplace partners provide ready-to-deploy, continuously updated protections for your cloud workloads. Managed rules for AWS Network Firewall are available from AWS Marketplace sellers of Check Point, Fortinet, Infoblox, Lumen, Rapid7, ThreatSTOP, and Trend Micro. AWS Marketplace rule groups are now available in 9 additional AWS Regions: Asia Pacific (Jakarta), Asia Pacific (Hyderabad), Asia Pacific (Melbourne), Asia Pacific (Malaysia), Canada West (Calgary), Europe (Zurich), Europe (Spain), Israel (Tel Aviv), and Mexico (Central). For a full list of supported regions, visit the AWS Regional Services page. To get started, visit the AWS Network Firewall console or browse available managed rules in AWS Marketplace. For more information, see the AWS Network Firewall product page and the service documentation.
Quelle: aws.amazon.com

Amazon Athena Spark adds support for AWS PrivateLink

Amazon Athena Spark now supports AWS PrivateLink so that you can access APIs and endpoints from your Amazon Virtual Private Cloud (VPC) without traversing the public internet. This feature can help you meet compliance requirements by allowing you to access and use Athena Spark APIs and endpoints entirely within the AWS network. You can now create AWS PrivateLink interface endpoints to connect from clients in your VPC. The Athena VPC endpoint supports all Athena Spark APIs and endpoints, including the Spark Connect, Spark Live UI and Spark History Server endpoints. Communication between your VPC and Athena Spark APIs and endpoints is then conducted entirely within the AWS network, providing a secure pathway for your data. To get started, you can create an interface VPC endpoint to connect to Amazon Athena Spark using the AWS Management Console or AWS Command Line Interface (AWS CLI) commands or AWS CloudFormation. This new feature is available in all AWS Regions where Amazon Athena Spark and AWS PrivateLink are available. For more information, refer to the AWS PrivateLink documentation and Athena Spark documentation.
 
Quelle: aws.amazon.com

Introducing the Amazon EKS Hybrid Nodes gateway for hybrid Kubernetes networking

Amazon Elastic Kubernetes Service (EKS) now offers the Amazon EKS Hybrid Nodes gateway, a feature that automates networking between your Amazon EKS cluster VPC and Kubernetes Pods running on Amazon EKS Hybrid Nodes. The Amazon EKS Hybrid Nodes gateway eliminates the need to make on-premises pod networks routable or coordinate network infrastructure changes when running in hybrid Kubernetes environments. Networking in hybrid Kubernetes environments can be complex, often requiring changes to on-premises routing configurations, coordination with network teams, and ongoing maintenance as workloads scale. The Amazon EKS Hybrid Nodes gateway addresses these challenges by automatically enabling Kubernetes control plane-to-webhook communication, pod-to-pod traffic across cloud and on-premises environments, and connectivity for AWS services such as Application Load Balancers, Network Load Balancers, and Amazon Managed Service for Prometheus. Customers deploy the Amazon EKS Hybrid Nodes gateway to Amazon EC2 instances using Helm, and the gateway automatically maintains VPC route tables as workloads scale. The Amazon EKS Hybrid Nodes gateway codebase is open source. The Amazon EKS Hybrid Nodes gateway is available in all AWS Regions where Amazon EKS Hybrid Nodes is available, except the China Regions. The Amazon EKS Hybrid Nodes gateway is offered at no additional charge. You pay for the underlying AWS infrastructure used to run the gateway, including Amazon EC2 instance charges and any associated data transfer fees. To get started, visit the Amazon EKS Hybrid Nodes gateway documentation.
Quelle: aws.amazon.com

Five new Qwen models for coding agents and efficient reasoning are now available in Amazon SageMaker JumpStart

Today, AWS announced the availability of Qwen3-Coder-Next, Qwen3-30B-A3B, Qwen3-30B-A3B-Thinking-2507, Qwen3-Coder-30B-A3B-Instruct, and Qwen3.5-4B in Amazon SageMaker JumpStart, expanding the portfolio of foundation models available to AWS customers. These five models from Qwen bring specialized capabilities spanning agentic coding, efficient reasoning, extended thinking, and multimodal understanding, enabling customers to build sophisticated AI applications across diverse use cases on AWS infrastructure.
These models address different enterprise AI challenges with specialized capabilities:
Qwen3-Coder-Next excels at long-horizon reasoning, complex tool use, and recovery from execution failures, making it ideal for powering coding agents in CLI/IDE platforms.
Qwen3-30B-A3B uniquely supports seamless switching between thinking and non-thinking modes, making it well suited for general-purpose assistant tasks like multilingual dialogue, math reasoning, and tool calling.
Qwen3-30B-A3B-Thinking-2507 delivers significantly improved performance on complex reasoning tasks in math, science, and coding, with enhanced long-context understanding.
Qwen3-Coder-30B-A3B-Instruct is designed for agentic coding workflows with a custom function call format and repo-scale context understanding.
Qwen3.5-4B supports unified vision-language training and  201 languages, making it ideal for lightweight multimodal deployments.
With SageMaker JumpStart, customers can deploy any of these models with just a few clicks to address their specific AI use cases.
To get started with these models, navigate to the Models section of SageMaker Studio or use the SageMaker Python SDK to deploy the models to your AWS account. For more information about deploying and using foundation models in SageMaker JumpStart, see the Amazon SageMaker JumpStart documentation.
Quelle: aws.amazon.com

Amazon SageMaker now supports multi-region replication from IAM Identity Center

Amazon SageMaker now supports multi-region replication from IAM Identity Center (IdC), enabling you to deploy SageMaker Unified Studio domains in different regions from your IdC instance. This new capability empowers enterprise customers, particularly those in regulated industries like financial services and healthcare, to maintain compliance while leveraging centralized workforce identity management. As an Amazon SageMaker Unified Studio administrator, you can deploy SageMaker domains closer to your workforce based on data residency needs while maintaining seamless single sign-on (SSO) access. Organizations can address use cases such as maintaining IdC in one region while processing sensitive data in compliance-required regions, supporting global operations with centralized identity management, and meeting data sovereignty requirements without compromising SSO capabilities.
To get started see the SageMaker Unified Studio documentation and to learn about setting up IAM Identity Center multi-Region support see the IAM Identity Center User Guide.
Quelle: aws.amazon.com

AWS Backup adds Amazon Redshift Serverless and Aurora DSQL support for AWS Organizations backup policies

AWS Backup now supports Amazon Redshift Serverless namespaces and Amazon Aurora DSQL clusters as resource types in AWS Organizations backup policies. Organization administrators can now define backup policy rules that directly target these resource types across member accounts.
Previously, backing up Redshift Serverless namespaces and Aurora DSQL clusters through organization backup policies required using tag-based selections or backing up all resources in a member account. With this launch, administrators can specify these resource types directly in their backup policy selections, providing more precise control over which resources are included in or excluded from Organization-wide backup plans.
This capability is available in all AWS Commercial and GovCloud Regions where AWS Backup and the respective services are available. To get started, visit the AWS Organizations backup policies documentation or the AWS Backup console.
Quelle: aws.amazon.com

Amazon Aurora serverless: Up to 30% better performance, smarter scaling, and still scales to zero

Amazon Aurora serverless — the autoscaling database that scales up to support your most demanding workloads and down to zero when you don’t need it — just got faster and smarter, with up to 30% better performance than the previous version and enhanced scaling that understands your workload. It’s especially well-suited for agentic AI applications, which typically have bursts of activity, long idle windows, and unpredictable patterns. Aurora serverless handles all of it automatically, scaling capacity with your agents rather than against them, and you only pay for what you actually use. When not in use, the database automatically scales down to zero to save cost. With improved performance and scaling, you can now use serverless for even more demanding workloads. The enhanced scaling algorithm enables you to efficiently run workloads where multiple tasks compete for resources, such as busy web applications and API services. These improvements are available in platform version 4 at no additional cost. All new clusters, database restores, and new clones will automatically launch on platform version 4. Existing clusters on platform version 1, 2, or 3 can upgrade directly to platform version 4 by using pending maintenance action, stopping and restarting the cluster, or using blue/green deployments. You can verify your cluster’s platform version in the AWS Console under instance configuration section or via the RDS API’s ServerlessV2PlatformVersion parameter. To learn more, read the blog. Aurora serverless is an on-demand, automatic scaling configuration for Amazon Aurora. For pricing details and Region availability, visit Amazon Aurora Pricing. To learn more, read the documentation, and get started by creating an Aurora serverless database using only a few steps in the AWS Management Console.
Quelle: aws.amazon.com

Amazon EC2 G7e instances now available in AWS Local Zones in Los Angeles

Today, AWS announces the general availability of Amazon Elastic Compute Cloud (Amazon EC2) G7e instances in AWS Local Zones in Los Angeles, California. G7e instances feature NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs and 5th generation Intel Xeon Scalable (Emerald Rapids) processors, bringing high-performance GPU compute closer to end users in Los Angeles. 
For creative workloads, you can use G7e instances to run studio workstation workloads with low-latency access to local storage, and post-production workloads including visual effects (VFX) editorial, color correction, and VFX finishing. G7e instances support enhanced real-time rendering on graphics engines and 2D/3D VFX composition software. For AI workloads, you can also use G7e instances to deploy Large Language Models (LLMs), inference, and agentic AI at the edge. 
To get started, opt-in to the Los Angeles Local Zone (us-west-2-lax-1b) from AWS Global View. You can enable G7e instances from the Amazon EC2 console, AWS Command Line Interface (AWS CLI), and AWS SDKs. G7e instances are available through On Demand and Savings Plans. To learn more, visit the AWS Local Zones Features page.
Quelle: aws.amazon.com

AWS Lambda functions can now mount Amazon S3 buckets as file systems with S3 Files

AWS Lambda now supports Amazon S3 Files, enabling your Lambda functions to mount Amazon S3 buckets as file systems and perform standard file operations without downloading data for processing. Built using Amazon EFS, S3 Files gives you the performance and simplicity of a file system with the scalability, durability, and cost-effectiveness of S3. Multiple Lambda functions can connect to the same S3 Files file system simultaneously, sharing data through a common workspace without building custom synchronization logic.
The S3 Files integration simplifies stateful workloads in Lambda by eliminating the overhead of downloading objects, uploading results, and managing ephemeral storage limits. This is particularly valuable for AI and machine learning workloads where agents need to persist memory and share state across pipeline steps. Lambda durable functions make these multi-step AI workflows possible by orchestrating parallel execution with automatic checkpointing. For example, an orchestrator function can clone a repository to a shared workspace while multiple agent functions analyze the code in parallel. The durable function handles checkpointing of execution state while S3 Files provides seamless data sharing across all steps.
To use S3 Files with Lambda, configure your function to mount an S3 bucket through the Lambda console, AWS CLI, AWS SDKs, AWS CloudFormation, or AWS Serverless Application Model (SAM). To learn more about how to use S3 Files with your Lambda function, visit the Lambda developer guide. 
S3 Files is supported for Lambda functions not configured with a capacity provider, in all AWS Regions where both Lambda and S3 Files are available, at no additional charge beyond standard Lambda and S3 pricing.
Quelle: aws.amazon.com