We have un-matched partnership networks with world-class tech vendors.
BeeGFS
BeeGFS is a parallel clustered file system developed with a strong focus on performance and designed for exceptionally easy installation and management. Originally created at the Fraunhofer Institute (as FhGFS) and now developed by ThinkParQ, BeeGFS transparently spreads user data across multiple servers; adding servers and disks scales both performance and capacity seamlessly, from small clusters to enterprise systems with thousands of nodes.
BeeGFS prides itself on simplicity without sacrificing parallel performance. It can be installed in hours, runs on commodity hardware on the top of standard local file systems, and requires minimal specialist administration. This makes genuine parallel I/O accessible to departmental clusters and mid-sized research facilities, not just national centres.
BeeGFS supports flexible building blocks with mirrored metadata, storage pools mixing NVMe and HDD tiers, and multi-rail RDMA networking (InfiniBand, RoCE). This allows OCF to design cost-effective systems that are matched to workload profiles.
AI training demands high-throughput streaming reads of millions of files; BeeGFS delivers strong small-file and metadata performance, and pairs effectively with Xinnor xiRAID for NVMe-backed GPU cluster storage.
BeeGFS' USP is providing parallel file system performance with the lowest administrative burden in its class.
