Intel has unveiled a new product line that marries its Xeon chips with a customizable FPGA chip, and it's promising that this combination will mean dramatically better performance for cloud service providers and carriers.
Diane Bryant, senior vice president of Intel's data center group, made the announcement at the GigaOm Structure conference in San Francisco on Wednesday.
Intel has been working on the product for "about a year," Bryant said. She declined to say when it will ship or to name the Intel customers that are currently testing it out.
Intel isn't disclosing the manufacturer of the FPGA (field-programmable gate array) chip.
Customers with apps that benefit from acceleration can see a 10-time to 20-time performance boost, Bryant said at the event. She described the Xeon-FPGA chip hybrid as "truly dynamic," adding that it will give large-scale customers more "agility and flexibility."
Intel began working with cloud service providers in 2007 on technology that lets data centers run more efficiently.
"We have been engaging directly with large-scale service providers to give them exactly what they need," Bryant said. "This has been an evolving, continuous process."
Asked if this is a move to stave off competition from vendors of servers running low-power ARM chips, Bryant suggested that this was one of the chip maker's motivations, but not the only one.
"We will always have competition; we're not naïve to that," Bryant said. "We are extremely paranoid about competition. We have invested heavily in adding product lines to make sure compute, storage and networking workloads run well on Intel."
Bryant acknowledged that customers are hesitant to rely on a single source for their IT needs. However, using multiple different chip vendors adds complexity and cost to data center management, she said.
Intel, with its new product line, is setting itself up as a one-stop shop for processors for its large-scale customers.
PUBLISHED JUNE 18, 2014