Hitachi Data Systems
Discover and Analyze Data for Competitive Advantage

With the movement of big data and advanced analytics from their traditional uses in university research to commercial businesses, high-performance computing (HPC) systems have now become ubiquitous. Additionally, as data recording is moving from traditional approaches on magnetic media to in-memory computing, which closely mirrors traditional HPC application processing. These movements have led to new requirements for data protection, cost-efficient infrastructures, east-west networks (server to server traffic that closely mirrors HPC fabrics) and reduced management workloads for administrators. Hitachi Data Systems (HDS), headquartered in California, uses HPC ideas such as building converged stacks which blend application, OS, VM, compute, network fabric and storage into a single package with a consistent management interface. “These systems address many of the concerns and pain points for users in the market. The management interface masks much of the complexity for managing single atomic elements and makes it simple to set up data protection schemes on a single site or multiple sites,” said Michael Hay, Vice President, Chief Engineer, Hitachi Data Systems.

HDS was founded as a global provider of storage solutions for commercial uses. Through the years, the company’s portfolio has expanded through compute and novel software capabilities for traditional commercial workloads to include the era of advanced analytics. “For the last several years in particular, we have focused on converged systems that support these HPC-like infrastructure-hungry applications,” said Hay. “Ultimately our aim is to build solutions for infrastructure, content and analytics for our vision of Social Innovation. Our goal for IT infrastructure and software platforms is to enable complex workloads and simulations for Social Innovation.” To achieve this, HDS combines sophisticated algorithms with advanced data processing infrastructures in centralized data centers.

The growth of HDS from basic storage supplier to complete systems supplier is best illustrated though one of the company’s flagship products, Hitachi Unified Compute Platform (UCP).

This system makes it possible to start with a storage capability, then upgrade the system in place with east-west networking and hundreds of compute blades to realize a complete converged system that meets the demands of data processing. “This ensures that our customers can grow from basic storage infrastructure supplier to HPC or converged systems operator over time without leaving their staff behind,” said Hay.

The company’s platforms are also coupled to a storage and systems eco¬nomics practice that saves money for IT customers. For instance, HDS analysis helped a Fortune 500 fi¬nancial services company address its rapidly rising costs of manag¬ing hundreds of remote servers by choosing a server consolidation solution. Impressed by the performance, scalability, and ease of SAN integration and management, the customer integrated the full portfolio of Hitachi Data Systems file and content solutions to consolidate its existing file servers onto fewer nodes. “This consolidation resulted in lower overhead, reduced footprint, and reduc¬tions in power and cooling costs that saved the client $5 million over the first three years,” said Hay.

Ultimately our aim is to build solutions for infrastructure, content and analytics for our vision of Social Innovation

For the future, HDS wants to build a complete platform that offers scale-up in-memory computation, intense scale-out for workloads, and edge computing for distant but latency-intolerant applications. “The Hitachi aim is to address all of these archetypes and the evidence comes initially in the form of our Hitachi Unified Compute Platform,” said Hay.