Hewlett Packard (HP) sees a future in super-efficient data centers powered by the kinds of power-sipping computer chips used in smartphones and tablet computers.
The computer giant was launched in Project Moonshot, server technology that cuts complexity, energy use, and costs, according to a description at the California-based company's website on Wednesday.
"It paves the way to the future of low-energy computing for emerging web, cloud and massive scale environments," HP said of Moonshot.
As people and businesses switch to using software as services hosted in the Internet "cloud" demand is rising for data centers with tremendous or "hyperscale" capacity to handle the tsunami of digital information.
"The volume of data processed in financial markets has increased exponentially, and traditional scale-up or scale-out architectures are struggling to keep up with demand without vastly increasing cost and power usage," said Cantor Fitzgerald high frequency trading director Niall Dalton.
"HP is taking a holistic approach to solving this problem and working to bring unprecedented energy and cost savings for tomorrow's large-scale, data-intensive applications," he continued.
The global financial services firm is evaluating the HP technology.
HP making Redstone servers built with the kinds of ARM chips used in mobile gadgets and combined with "Converged Infrastructure" technology that spreads word loads over thousands of machines.
"Companies with hyperscale environments are facing a crisis in capacity that requires a fundamental change at the architectural level," said HP hyperscale business unit vice president Paul Santeler.
Moonshot provides a server development platform along with a lab for customers and partners.
HP expected its new data center approach to reduce energy consumption for some types of workloads by as much as 89 percent while requiring less space for equipment and cutting costs up to 63 percent.
Critics question wether mobile gadget chips have the strength for data center needs, noting that software must be carefully crafted to divvy up workloads or risk overwhelming processors.
Explore further: Intel has end-to-end reference model for IoT