Case Study: Siemens Healthineers achieves scalable AI Factory storage with Quobyte

How Siemens Healthineers manages storage for its AI Factory infrastructure using Quobyte

Siemens Healthineers’ Digital Technology and Innovation team needed a production-ready, scalable storage platform for its AI Factory and Sherlock supercomputer as deep learning datasets and experiment volumes grew rapidly. After evaluating other parallel file systems, the team turned to Quobyte because it could run on existing commodity hardware, support direct high-speed access for multi-terabyte AI training jobs, and provide the reliability, fault tolerance, and security required for a global healthcare R&D environment.

With Quobyte, Siemens Healthineers standardized on a unified scale-out storage system with 3-way replication, erasure coding, automatic tiering, ACLs, S3 compatibility, and seamless NFS/POSIX access. The result was lower operating costs, easier administration, no-downtime capacity expansion, and stronger data governance and business continuity. Quobyte also supported the migration of existing storage one server at a time without disrupting production, while enabling massive AI workloads such as training on 100 million to 500 million medical images and continuing to scale for large language models.

Siemens Healthineers

Gianluca Paladini

Sr. Director of Engineering