DataDirect Networks (DDN) has announced a new version of its high-performance Exascaler storage array.

Exascaler 2.1 offers more IOPS than its predecessor and brings a host of software improvements, helping it scale to tens of thousands of clients and petabytes of storage. The array is aimed at enterprise high performance computing (HPC) environments in fields such as oil and gas, financial services and manufacturing.

The hardware relies on Intel’s distribution of the open source Lustre file system. DDN says it has expanded its relationship with Intel, a company credited with taking Lustre mainstream.

Luxury storage
DDN is the world's largest privately owned storage company, especially prominent in the HPC market. Among other things, it built the storage system for Titan – currently the world’s second most powerful supercomputer.

Exascaler is the company’s flagship array, integrating file system, RAID controllers and storage media into a single, massive appliance built with DDN’s Storage Fusion Architecture (SFA). The hardware runs Intel Enterprise Edition for Lustre, a set of tools for the parallel distributed file system that excels at moving data at petabyte scale.

Exascaler 2.1 can squeeze 1680 drives into just two racks, to provide up to 40GB/s bandwidth and up to 1.4 million IOPS. The latest version of the array adds improved drive rebuild times, self-encrypting drive support and enhanced remote diagnostics tools.

It also features a unique read cache technology called Software Fusion Xcelerator (SFX) to improve metadata performance, especially for operations with small files.

Exascaler owners can use a software adapter to run MapReduce applications directly on top of Lustre, speeding up analytics applications.

DDN says that all of these features help the array deliver around 25 percent more performance than competing Lustre solutions.

“A massively scalable, efficient appliance designed for the HPC and Big Data markets, EXAScaler delivers extreme file system performance to help simplify and make it more affordable for customers to build petascale computing solutions for data hungry applications across sectors including scientific research, trading simulation, climate modeling and energy exploration,” said Michael Vildibill, vice president of Product Management at DDN.