NumaRAID is our fully parallel RAID stack that we designed from the outset to exploit the power of multi-core and many-core (MIC) processors. NumaRAID is the culmination of over 20 years of experience in RAID architectures, and was designed specifically for reliability at scale. It offers tremendous bandwidth and processing power and supports both multi-processor and multi-memory motherboards. With direct support for both multi-port FDR Infiniband and multi-port Fibre Channel, NumaRAID has been measured at rates exceeding 4 Gigabytes/Second in single node configurations, even with two failed drives. >>Link to NumaRAID product

NumaRAID supports both a Command Line Interface (CLI) and a Graphical User Interface (GUI). NumaRAID provides full monitoring of drives, power supplies, fans and temperature readings. It includes direct support for Email notifications, SNMP traps, and has extensive performance measurement tools that are both easy to use and extremely detailed. It has been qualified by over 20 independent companies. Under different OEM brands, NumaRAID is a leading product in the Media and Entertainment, High Performance Computing and Health Care markets. >>NumaRAID product deployment

NumaRAID has been field deployed for over 4 years, and has thousands of active installations around the world. Since NumaRAID is hardware independent, each new generation of X86 processor from Intel or AMD further accelerates its performance advantage over proprietary hardware solutions. The latest generation of NumaRAID supports PCI-E 3.0 and the Intel Xeon E5 processors and has a single node performance target of 10 Gigabytes/Second.

NumaRAID executes in a RHEL5.2 environment, and is compatible with StorNext, MetaSan and HyperFS, which eliminates the need for a separate hardware Metadata Server. NumaRAID can also be configured as a Lustre Object Storage Server node, eliminating the need for a separate OSS server.

NumaRAID is the only fully supported high performance commercial RAID system available with full source code. Contact StreamScale for more information.