vePFS supports RDMA and delivers bandwidth of up to one terabyte per second, tens of millions of IOPS, and sub-millisecond latency. Its performance scales linearly with capacity, providing exceptional storage performance.

High scalability

vePFS supports online elastic expansion, enabling you to quickly increase capacity and achieve linear growth in both capacity and throughput performance at any time.

Stable and easy to use

vePFS is based on a multi-level reliability architecture that ensures efficient and stable operation of the file systems. Boasting a data durability of 99.9999999% (nine nines), it offers a user-friendly web interface. Users can manage resources via the console or API.

Targeted optimization for model training

vePFS is optimized in conjunction with the machine learning platform (AML) and GPU computing resources to provide high-quality storage services for training scenarios.

Last updated: 2026.03.03 10:17:07

Public security network record in Beijing: No. 11010802032137