
Weka
WEKA’s high-performance storage software runs natively on Nebius hardware, integrated into the Nebius AI Cloud stack. Customers running large training and inference workloads get microsecond-latency data access at petabyte-to-exabyte scale — storage that keeps pace with GPU clusters.
Announced in 2025, the partnership serves research and enterprise deployments, including multi-petabyte research workloads.
See WEKA announcement: WEKA and Nebius Partner to Deliver High-Performance AI Data Infrastructure.