Amazon Web Services (AWS) announced the general availability of S3 Files, a new capability that enables a shared file system directly on top of Amazon S3, delivering high-performance file access without moving data out of object storage.
With S3 Files, Amazon S3 is the first cloud object store to provide fully featured file system semantics alongside low-latency performance. The solution allows file-based applications, AI agents, and teams to access and work with S3 data as a file system using existing tools, without requiring code changes or data duplication.
Built using Amazon Elastic File System (EFS), S3 Files combines the simplicity and performance of a file system with the scalability, durability, and cost efficiency of S3. The service maintains a consistent view of objects within S3 buckets and automatically translates file system operations into optimized S3 requests, eliminating the need for complex data pipelines or synchronization processes.
S3 Files addresses a longstanding challenge for organizations that store analytics data and data lakes in S3 but rely on file-based tools and applications. By unifying file and object storage, the new capability removes the need to manage separate systems or duplicate datasets, enabling seamless access to data through both file system interfaces and S3 APIs simultaneously.
The solution supports thousands of concurrent connections from compute resources, including instances, containers, and serverless functions, enabling shared access across large-scale environments. It also features intelligent caching for frequently accessed data, delivering low-latency performance and supporting multiple terabytes per second of aggregate read throughput.
S3 Files is compatible with existing and new data in S3 buckets, requiring no migration. The service is now generally available in 34 AWS Regions, providing organizations with a streamlined approach to managing and accessing data at scale.
