All news with #sagemaker tag

AWS launches Apache Spark Upgrade Agent for Amazon EMR

🛠️ AWS announced the Apache Spark upgrade agent, a capability that automates and accelerates Spark version upgrades for Amazon EMR on EC2 and EMR Serverless. The agent performs automated code analysis across PySpark and Scala, identifies API and behavioral changes for Spark 2.4→3.5, and suggests precise code transformations. Engineers can invoke the agent from SageMaker Unified Studio, the Kiro CLI, or any MCP-compatible IDE, interact via natural-language prompts, review proposed edits, and approve implementations. Functional correctness is validated through data quality checks to help maintain processing accuracy during migration.

Amazon Nova Forge: Build Frontier Models with Nova

🚀 Amazon Web Services announced general availability of Nova Forge, a SageMaker AI service that enables organizations to build custom frontier models from Nova checkpoints across pre-, mid-, and post-training phases. Developers can blend proprietary data with Amazon-curated datasets, run Reinforcement Fine Tuning (RFT) with in-environment reward functions, and apply custom safety guardrails via a built-in responsible AI toolkit. Nova Forge includes early access to Nova 2 Pro and Nova 2 Omni and is available today in US East (N. Virginia).

Amazon SageMaker Catalog Adds Automated Data Classification

🤖 Amazon SageMaker Catalog now provides automated data classification that suggests business glossary terms during dataset publishing to reduce manual tagging and improve metadata consistency. The capability leverages Amazon Bedrock language models to analyze table metadata and schema and recommend relevant business and sensitive-data terms from organizational glossaries. Data producers receive AI-generated suggestions they can accept or modify before publishing, helping standardize vocabulary and improve data discoverability. The feature is available in multiple AWS regions and can be managed via SageMaker Unified Studio, the AWS CLI, or SDKs.

SageMaker AI Adds Flexible Training Plans for Inference

⚙️ Amazon SageMaker AI's Flexible Training Plans (FTP) now support inference endpoints, allowing customers to reserve guaranteed GPU capacity for planned evaluations and production peaks. You choose instance types, compute requirements, reservation length, and start date, then reference the reservation ARN when creating an endpoint. SageMaker AI automatically provisions and runs the endpoint on the reserved capacity for the plan duration, removing much of the infrastructure scheduling overhead. FTP for inference is initially available in US East (N. Virginia), US West (Oregon), and US East (Ohio).

Manage SageMaker HyperPod Clusters with AI MCP Server

🔧 The Amazon SageMaker AI MCP Server now provides tools to set up and manage HyperPod clusters, allowing AI coding assistants to provision and operate clusters for distributed training, fine‑tuning, and deployment. It automates prerequisites and orchestrates clusters via Amazon EKS or Slurm with CloudFormation templates that optimize networking, storage, and compute. The server also delivers lifecycle operations — scaling, patching, diagnostics — so administrators and data scientists can manage large-scale AI/ML clusters without deep infrastructure expertise.

Amazon SageMaker Adds EAGLE for Faster Inference Throughput

⚡ Amazon SageMaker AI now supports EAGLE (Extrapolation Algorithm for Greater Language-model Efficiency) speculative decoding to boost large language model inference throughput by up to 2.5x. The capability enables models to predict and validate multiple tokens in parallel rather than one at a time, preserving output quality while reducing latency. SageMaker automatically selects between EAGLE 2 and EAGLE 3 depending on model architecture and provides built‑in optimization jobs using curated or customer datasets. Optimized models can be deployed through existing SageMaker inference workflows without infrastructure changes, and the feature is available in select AWS Regions.

SageMaker HyperPod Adds NVIDIA MIG GPU Partitioning

🚀 Amazon SageMaker HyperPod now supports NVIDIA Multi-Instance GPU (MIG), enabling administrators to partition a single GPU into multiple isolated devices to run simultaneous small generative AI tasks. Administrators can use an easy console configuration or a custom setup for fine-grained hardware isolation, allocate compute quotas across teams, and monitor real-time performance per partition via a utilization dashboard. Available on HyperPod clusters using the EKS orchestrator in multiple AWS Regions, this capability reduces wait times by letting data scientists run lightweight inference and interactive notebooks in parallel without consuming full GPU capacity.

Amazon SageMaker Unified Studio Adds Catalog Notifications

🔔 Amazon SageMaker Unified Studio now delivers real-time notifications for data catalog activities, including new dataset publications, metadata changes, subscription requests, comments, and access approvals. Alerts are surfaced via a bell icon on the project home page and through a notification center that shows a recent list and a full, filterable tabular view by catalog, project, and event type. The feature is available in all regions where SageMaker Unified Studio is supported.

Aurora PostgreSQL zero-ETL now integrates SageMaker

🔁 Amazon Aurora PostgreSQL-Compatible Edition now offers zero-ETL integration with Amazon SageMaker, enabling near-real-time replication of PostgreSQL tables into a lakehouse. The synced data conforms to Apache Iceberg open standards and is immediately accessible to SQL, Apache Spark, BI, and ML tools via a simple no-code interface without impacting production workloads. Comprehensive, fine-grained access controls are enforced across analytics engines, and the capability is available in multiple AWS Regions.

SageMaker AI Projects Adds Custom ML Templates from S3

🛠️ Amazon Web Services announced that SageMaker AI Projects can now provision custom ML project templates stored in Amazon S3. Administrators can define and manage standardized end-to-end project templates in SageMaker AI Studio so data scientists can create projects that follow organizational patterns and automated workflows. The feature is available in all AWS Regions where SageMaker AI Projects is offered.

SageMaker Unified Studio adds SSO for Spark sessions

🔐 Amazon SageMaker Unified Studio now supports corporate identities for interactive Apache Spark sessions using AWS Identity Center trusted identity propagation. Data engineers and scientists can sign on to JupyterLab Spark sessions with organizational credentials while administrators apply fine-grained access controls and maintain end-to-end data access traceability. The integration leverages AWS Lake Formation, Amazon S3 Access Grants, and Amazon Redshift Data APIs, and includes comprehensive AWS CloudTrail logging for interactive and background sessions to streamline compliance.

Amazon SageMaker HyperPod Adds Managed Karpenter Autoscaling

🛠️ Amazon SageMaker HyperPod now supports managed node autoscaling using Karpenter, enabling automated cluster scaling for both inference and training workloads. This managed capability removes the operational burden of installing and maintaining autoscaling infrastructure while providing integrated resilience and fault tolerance. Customers gain just-in-time GPU provisioning, scale-to-zero during low demand, workload-aware instance selection, and cost reductions through intelligent consolidation.

Amazon SageMaker Adds EC2 P6-B200 Notebook Instances

🚀 Amazon Web Services announced general availability of EC2 P6-B200 instances for SageMaker notebooks. These instances include eight NVIDIA Blackwell GPUs with 1,440 GB of high-bandwidth GPU memory and 5th Gen Intel Xeon processors, offering up to 2x the training performance versus P5en. They enable interactive development and fine-tuning of large foundation models in JupyterLab and CodeEditor, and are available in US East (Ohio) and US West (Oregon).

Amazon SageMaker Adds Restricted Classification Terms

🔒 Amazon SageMaker Catalog now supports governed classification using Restricted Classification Terms, enabling catalog administrators to mark sensitive glossary terms so only authorized users or projects can apply them to assets. Administrators grant usage through explicit policies and group membership, allowing centralized governance teams to control labels like Seller-MCF or PII. The capability is available in all regions that support SageMaker Unified Studio; consult the user guide to get started.

AWS SageMaker Adds P5.4xlarge with NVIDIA H100 GPU

🚀 Amazon SageMaker Training and Processing Jobs now supports the new EC2 P5 instance size with a single NVIDIA H100 GPU, offering the P5.4xlarge configuration for cost‑effective ML and HPC workloads. The instance enables fine-grained scaling so customers can begin with smaller configurations and expand incrementally, improving cost management and infrastructure flexibility. P5.4xlarge is available via SageMaker Flexible Training Plans and in select regions through On‑Demand and Spot.