Ddos 
At re:Invent 2025, AWS announced a sweeping set of enhancements to its flagship storage service, Amazon S3, designed expressly for the era of AI and big data. Chief among them is the launch of the general availability of Amazon S3 Vectors, which introduces native vector storage and querying capabilities, as well as a dramatic expansion of the maximum Amazon S3 object size to 50 TB, empowering enterprises to build generative-AI applications and data lakes more efficiently and at lower cost.
To enable AI systems to store and query vector data directly within the native S3 environment, AWS officially released Amazon S3 Vectors GA.
Compared with the earlier preview release, the production version delivers a substantial leap in scale: a single index now supports up to 2 billion vectors (a 40× increase in capacity), and a single bucket can store up to 20 trillion vectors.
AWS emphasizes that Amazon S3 Vectors delivers 2–3× faster performance for high-frequency queries and enables customers to reduce infrastructure spending by up to 90%, while eliminating the operational burden of maintaining dedicated vector-database systems.
The feature is deeply integrated with Amazon Bedrock Knowledge Bases and Amazon OpenSearch Service, allowing organizations to more easily build AI agents and retrieval-augmented-generation (RAG) systems. Early adopters include BMW Group, mixi, and Twilio.
As global data volumes continue to soar, AWS raised the maximum S3 object size from 5 TB to 50 TB, a tenfold increase. This enables customers to store massive individual files without splitting them — such as ultra-high-resolution video, large-scale seismic-survey data, and enormous AI training datasets — while retaining all S3 storage classes and features.
Amazon S3 Batch Operations also received a major performance upgrade. Large-scale batch workflows — such as cross-region replication, tag management, or checksum verification — can now run up to 10× faster, with each job supporting up to 20 billion objects.
For Apache Iceberg workloads, Amazon S3 Tables added two key capabilities:
- Intelligent-Tiering support: automatically moves data between three access tiers according to usage patterns, cutting storage costs by up to 80% without compromising performance.
- Automatic Replication: enables cross-region and cross-account table replication, improving global data access while simplifying compliance and backup operations.
Finally, AWS expanded Amazon S3 Access Points to support Amazon FSx for NetApp ONTAP, allowing customers to access FSx-hosted files with the same ease as S3 objects. Enterprises can now apply S3-based tools — such as Amazon SageMaker, Athena, or Bedrock — directly to datasets stored in FSx, dramatically reducing the complexity of modernizing on-prem NetApp environments for cloud-native analytics and AI.
Beyond storage, AWS announced numerous updates across compute, databases, and modernization tools:
Compute Instances & Serverless Evolution
- Amazon EC2 X8aedz instances: powered by 5th-generation AMD EPYC processors offering up to 5 GHz clock speeds and 3 TB RAM, ideal for EDA workloads and memory-intensive databases.
- AWS Lambda Managed Instances: combines serverless simplicity with EC2 flexibility, enabling Lambda functions to run on EC2 hardware for easier access to specialized compute and superior cost optimization.
- AWS Lambda Durable Functions: supports multi-step, long-running workflows that can persist from seconds to an entire year, without incurring idle-compute charges while waiting for external events.
Databases & Search Optimization
- Database Savings Plans: a new pricing model offering greater flexibility and cost efficiency for AWS database services.
- Amazon OpenSearch Service acceleration: introduces GPU-powered indexing and automated optimization — achieving 10× faster large-scale vector-database construction at one-quarter the cost.
- RDS for SQL Server / Oracle: improved scalability and cost efficiency, including support for SQL Server Developer Edition and Oracle M7i/R7i instances.
Application Modernization & Management
- AWS Transform: adds Custom mode, using AI to learn organizational patterns and automate large-scale code transformation; introduces full Windows and Mainframe modernization capabilities to eliminate technical debt.
- Amazon EKS Capabilities: simplifies containerized-application development with fully managed workload orchestration and cloud-resource management.
- Amazon CloudWatch: introduces unified data management and analytics with support for OCSF and Apache Iceberg formats.
Networking & Partner Ecosystem
- Amazon Route 53 Global Resolver (preview): offers unified hybrid DNS management with Anycast-based global resolution for public and private domains.
- AWS Partner Central: now integrated directly into the AWS Management Console, streamlining partner-solution discovery and management.
Donate