The next leap in compute and storage density.

 

An NVMe converged appliance that combines Server, Storage & Network in a fully redundant 2U configuration.

 

Better Hardware: Our system utilizes dual ported NVMe Flash drives. NVMe Flash storage is 6X faster than traditional SAS based Flash. Using Intel hardware allows us to adapt quickly to the latest technology offerings. 

 

Better Architecture: The NVMe switching architecture, FabricExpress, was designed from the ground up to deliver ultra-high performance between the server modules and NVMe storage. The non-transparent bridge architecture that communicates between the server modules has less than 5 microseconds (uS) of latency.

 

Breathtaking performance:

  • 12 Million IOPs (Input/output operations per second)

  • 50 microseconds of latency

  • 480Gbps throughput

  • 1PB onboard storage

  • 2 U footprint

  • 4000 watts

 

Better Strategy:

  • Active/Active write acknowledgments are completed over the NVMe Fabric.

  • The System can also serve storage to external Hosts at up to 100GbE speeds.

 

Better Features:

  • Deduplication & Compression for Maximum Space Efficiency & Industry Leading Performance.

  • Scale In First - Data Locality for Ultra-Fast Local Write Acknowledgements via NVMe Fabric

  • Active/Active writes & Failover via Non-Transparent Bridge (NTB) 5 uS Response Times

  • Then Scale Out- Asynchronous Replication for Backup & D/R

  • Built for Microsoft Server 2016 &2016 Hyper V

  • Compatible with VMware and KVM

  • 50 % Reduction in Software Core-Based Licensing Costs – Axellio doubles CPU & RAM efficiency Better than Scale-out Traditional Scale Out Hyper-Converged Technologies:

  • Traditional Scale Out Technologies are up to 100X slower than Scale in 1st NVMe Technologies

  • Traditional Scale Out Technologies require double the CPU and RAM of Scale In 1st Technologies Better than Traditional Storage Arrays:

  • The cost for separate high-performance storage, networking, and servers is much higher than NVMe converged solutions in terms of both CAPEX and OPEX.

  • Traditional SANs (Even ALL Flash) are 6X to 100X slower than NVMe Converged Solutions.

  • Traditional SAN performance is dependent upon configuration of the storage, network, and hosts

  • Converged solutions are turn key and already optimized for performance.

 

Better Value:

Gain unparalleled performance and availability, while simplifying administration and reducing costs, in a 2U fully redundant, feature rich, ultra-converged solution.

 

  • Offers features including high availability, deduplication, and Asynchronous replication.

  • Delivers up to 100X better performance than the competition.

  • Reduces infrastructure costs by over 50%. Savings come from the convergence of storage, networking, and servers, as well as, the reduction in software licensing and in power, space, and cooling requirements.

NVMe Super Computer

NEXT GENERATION COMPUTE AND STORAGE

Optimizing your server infrastructure

This system delivers world record performance for demanding applications like cyber security, bioinformatics/medical research and big data analytics.