Fusion

Fusion, developed by Seqera, is a virtual, lightweight, distributed file system designed to bridge the gap between cloud-native data pipelines and cloud object storage. It is built for bioinformatics and life sciences teams running containerized workflows on cloud infrastructure, enabling seamless filesystem I/O to object stores such as AWS S3, Google Cloud Storage, and Azure — delivering over twice the throughput of standard cloud object storage access.

Traditional cloud pipelines face a fundamental challenge: object stores like S3 are scalable and cost-effective but do not present a POSIX interface. This forces containerized applications to copy data to and from storage for every task, creating slow, inefficient, and costly workflows. Fusion eliminates this bottleneck by providing direct, POSIX-compatible access to cloud object stores, simplifying pipeline logic and dramatically accelerating execution.

Key Benefits

• Simplify pipeline development and deployment by removing the need to manage data staging logic
• Reduce redundant file I/O and maximize compute resource use efficiency
• Avoid the need to pre-install cloud tools in containers or cloud instances
• Accelerate individual task and overall pipeline execution
• Eliminate the need for expensive and complex shared file systems
• Boost productivity and reduce time to results

Transparent, Automated Installation

• Traditionally, pipeline developers had to bundle utilities in containers to copy data in and out of S3 storage
• With Fusion, there is nothing to install or manage manually — the Fusion thin client is automatically installed and configured for each pipeline
• Containerized applications can read and write to S3 buckets, Google Cloud Storage, Azure, and other object stores as if they were local storage

Dramatically Reduced Data Movement

• Standard pipelines read data from a bucket, copy it to local block storage for processing, then copy results back — adding significant overhead for every task
• Fusion enables direct file access to object storage, eliminating unnecessary I/O and dramatically reducing data movement and overall runtime
• Fusion 2.0 outperforms existing object stores and file systems by avoiding intermediate data copies and leveraging a high-performance Fusion driver

No Shared File System Required

• Organizations typically rely on shared file systems such as Amazon EFS, Amazon FSx for Lustre, or NFS to share data among pipeline tasks
• Fusion removes the need to deploy, manage, and mount shared file systems on every cloud instance
• The same data-sharing functionality is delivered directly over cloud object stores such as S3, significantly reducing cost and operational complexity

Seamless Access to Cloud Object Storage

• Unlike third-party open-source POSIX-over-object-storage solutions that require manual installation and configuration, Fusion handles these tasks automatically
• Fusion is optimized specifically for Nextflow pipelines, delivering fast and seamless access without additional developer effort
• By eliminating the need for block storage and accessing object storage via a POSIX interface, efficiency is increased dramatically

Maximized Pipeline Performance and Cost Efficiency

• Copying data to and from object storage adds latency for every task, lengthening the time containers and cloud instances remain deployed
• For pipelines with thousands of tasks, this translates into significantly longer runtimes and higher cloud costs
• Fusion eliminates these bottlenecks, reducing execution time, lowering cloud spending, and using compute instances more efficiently

Fusion is purpose-built for cloud-native compute environments and integrates directly with Nextflow and the Seqera platform. It provides a simple, scalable file system solution that works across major cloud providers, removing infrastructure complexity and enabling teams to focus on scientific outcomes rather than data movement logistics.