HP has unveiled a line of its ProLiant server hardware designed for big data analysis.
The company said that its ProLiant SL4500 servers would match up compute nodes with storage and networking hardware to streamline the performance and reduce the cost of running big data analysis applications.
The systems, which start at $7,600 and are currently available for order, have been designed to reduce the physical footprint and interconnect requirements of using conventional server hardware for big data.
HP estimates that its new systems require 50 percent less space and 63 percent fewer interconnects than previous offerings.
Daniel Bounds, director of product management for HP, told V3 that the company was looking to cut back on the hardware costs and complexity commonly associated with launching a big data in the enterprise space.
"We offered the opportunity to our engineers about a year ago, if they had a clean slate to design a server for big data, what would they do," Bounds explained.
"Customers are able to take their number one line item on their operating budget and take a big chunk out of that."
The company said that the systems have been designed to fit into four to five units, combining compute nodes with up to 60 hard drives, depending on the intended use case and requirements of the customer.
In addition to meeting new hardware requirements, the company is also looking to accommodate new usage models for big data.
Customers have begun to demand better performance and faster results from analytics platforms as the big data market has grown, HP industry standard servers and software director for smart storage Jimmy Daley told V3.
"Hadoop started out in most instances as something you ran on the side and the result was put into whatever else you were doing," he said.
"Today, many people have their analytics tier as a real-time overlay of what is going on which, has driven performance requirements and data traffic requirements into a traditional data infrastructure."