IDF 2014: Intel’s Xeon D SoC For Microservers Enters Testing Stage

Max 'Beast from the East' Smolaks covers open source, public sector, startups and technology of the future at TechWeekEurope. If you find him looking lost on the streets of London, feed him coffee and sugar.

Follow on: Google +

Xeon D family built on Broadwell architecture is on track for release in 2015

Intel has begun testing its Xeon D processor family among select customers, with wider availability expected next year.

Xeon D is the company’s first System-on-a-Chip (SoC) under the Xeon brand. Built on the 14nm process, it enjoys low power consumption making it suitable for hyperscale environments, among other things.

The announcement was made by Intel’s general manager of the Data Centre Group Diane Bryant at the Intel Developers Forum (IDF) in San Francisco. “You get all the benefits of Xeon class reliability and performance, but you also get the very small footprint, high integration of the SoC capability,” she said.

Small but mighty

Xeon D is the third generation of Intel 64-bit SoCs for low power servers, this time based on ‘Broadwell’ architecture. It is aimed at hyperscale environments, low to mid-range networking and storage applications.

badge-xeon-vert.png.rendition.cq5dam.thumbnail.200.132The term ‘hypercale’ describes high density systems which contain a very large number of relatively underpowered processors, as opposed to the big, hot and power-hungry chips traditionally used in the data centre. Examples of such systems include ARM-based servers from Dell and HP’s recently launched Moonshot, which is built on Atom C2000 processors.

“The change can create a threat or it can create an opportunity. We obviously see it as a tremendous opportunity,” Bryant said.

“In the future, the data centre will be tuned into the workload, all infrastructure will be software-defined, and data analytics will be explicitly and implicitly part of all services.”

The advantage of SoCs is they can include things like networking components, Field-Programmable Gate Arrays (FPGAs) or hardware encryption directly on the processor die, in the same footprint, resulting in more capable, customisable silicon.

The new chips require as little as 15W power to operate, and share the Broadwell architecture with the Core M processors for PCs.

Core M can run on just 4.5W and Intel president Renee James previously referred to this family as the most energy efficient Core processors in Intel’s history. These chips will power thin, fanless devices including tablets, convertibles and All-in-Ones, going on sale as early as next month.

Bryant said the development of Xeon D was on track, and the chips would be hitting production in the first half of 2015.

Several days before, Intel announced a new Xeon product family designed to aid Software-Defined Infrastructure (SDI) deployments – the Xeon E5 v3, built on 22nm process using Haswell architecture.

What do you know about Intel? Take our quiz!