GPU-accelerated Block Device

  • Enables low cost high-density data storage

  • Create DAS arrays with hundreds of parallel storage devices

  • High performance even with massive degradation

  • GPU computing on storage nodes with minimal data movement

  • Compatible with all Linux filesystems and applications

NSULATE™ is available through our distribution partners. For a trial license, contact us.

Say Hello to NSULATE™

RAID was standardized in 1993 in an era of single-core computing. For exascale computing, RAID is obsolete and an obstacle to higher performance and resilience. NSULATE revolutionises the role of the storage controller by replacing a fixed-function RAID controller with a powerful general-purpose GPU. Using a GPU as a storage controller inserts an enormous amount of general purpose processing power directly into the storage pipeline, enabling modern HPC storage appliances to deliver unprecedented speed, scale, security, storage efficiency and intelligence in real-time.

Extreme Resilience

NSULATE offers extreme resilience. It uses a GPU to generate huge parity calculations to enable automatic data recovery on scales impossible with a RAID card or a CPU. As a Linux block device, it is compatible with existing Linux filesystems and applications.

While traditional RAID and erasure coding solutions support parity calculations up to 8, NSULATE supports up to 128 parity. Stable I/O throughput can be maintained even while experiencing dozens of simultaneous device failures and corruption events across an array.

Continuous Verification

NSULATE is compatible with Nyriad's NCRYPT™ suite of GPU-accelerated, NIST compliant cryptographic algorithms. NCRYPT includes real-time blockchain generation APIs that can apply any appropriate combination of encryption, hashing and cryptographic signature generation modules to all storage transactions. Integrated with NSULATE, NCRYPT enables all storage transactions to be encrypted, auditable for correctness/integrity and signed by any number of authenticated trust contracts. A Merkle DAG can be generated in real-time and linked to a private or public blockchain ledger for independent verification and proof of data integrity.

Converged Storage-Processing

NSULATE can further reduce infrastructure requirements by sharing GPU resources for compute and storage on the same physical node. Storage nodes can be configured to double as processing nodes for I/O bound computing steps. This further accelerates big data and HPC processing and storage access by reducing the distance between GPU resources and storage.

Use Case

NSULATE™ with Lustre® filesystem

NSULATE was originally designed for accelerating HPC storage for the Square Kilometer Array Telescope project which requires GPU processing and storage of over 50PB/day of astronomical data. It consequently pairs well with disributed and parallel HPC filesystems like Lustre, HDFS, and Ceph to significantly reduce the network and CPU overhead of storage-processing.

Lustre has traditionally suffered from challenges with reliability while maintaining performance. For ldiskfs-based Lustre configurations (EXT4), NSULATE adds high parity resilience and cryptographic checksum features. For ZFS-based configurations, NSULATE takes the burden of these calculations off the CPU freeing up resources for other ZFS capabilities.


NSULATEā„¢ is available to the market from our OEM partners.

If you are a system integrator interested in licensing NSULATE, contact [email protected].