Inspur AS13000G7 Distributed Storage System

The Inspur AS13000G7 is a seventh-generation software-defined, scale-out distributed storage system engineered to deliver unified block, file, and object storage services from a single converged platform. Built on a shared-nothing architecture, the AS13000G7 eliminates single points of failure through fully distributed metadata management and multi-copy or erasure-coding data protection schemes, enabling linear capacity and performance scaling as additional nodes are added to the cluster. The system is purpose-built for the demands of AI training pipelines, large-scale data lakes, and high-throughput analytics environments where concurrent access from thousands of clients is routine.

The AS13000G7 introduces enhanced NVMe-native acceleration paths and an upgraded internal network fabric compared to its predecessor generation, allowing it to sustain high IOPS and throughput workloads without bottlenecks at the storage layer. It supports multiple industry-standard protocols simultaneously—including NFS, SMB, S3-compatible object, iSCSI, and Fibre Channel—making it interoperable with heterogeneous compute clusters, GPU server farms, and existing enterprise infrastructure. Intelligent tiering capabilities allow hot data to reside on all-flash nodes while warm and cold data migrates automatically to high-density hybrid or capacity-optimized nodes within the same cluster namespace.

Designed for mission-critical enterprise and hyperscale-adjacent deployments, the AS13000G7 integrates with Inspur's InStorage management suite, providing a unified GUI and RESTful API surface for provisioning, monitoring, quality-of-service policy enforcement, and capacity planning. The platform supports deployment in on-premises data centers as well as private cloud environments and is a strong fit for organizations building AI data lakes, genomics repositories, media and entertainment post-production workflows, and petabyte-scale backup and archive tiers.

Ideal for

• AI and machine learning data lake storage, providing high-throughput parallel data ingestion and retrieval for GPU training clusters accessing large unstructured datasets concurrently
• Big-data analytics repositories where Hadoop, Spark, or similar frameworks require scalable, S3-compatible or HDFS-compatible object and file storage with low-latency metadata operations
• High-performance computing (HPC) scratch and persistent storage for scientific simulations, genomics sequencing pipelines, and computational fluid dynamics workloads demanding sustained throughput
• Media and entertainment digital asset management and post-production workflows requiring simultaneous high-bandwidth access to large video files from editing suites and rendering farms
• Enterprise backup, long-term archival, and compliance data retention at petabyte scale, leveraging erasure coding for space-efficient protection without sacrificing data durability
• Private cloud object storage infrastructure serving S3-compatible API endpoints for internal application teams and DevOps pipelines requiring elastic, self-service storage provisioning

Technical specifications

Available from Omnixon Global. Submit an RFQ and our team will confirm configuration and availability for your order.