Welcome!

SOA & WOA Authors: Michelle Drolet, Richard Moulds, Jim Liddle, Rakesh Shah, David Dodd

Related Topics: Cloud Expo, Java, SOA & WOA, Virtualization, Web 2.0, SDN Journal

Cloud Expo: Article

Intel Re-Imagines the Rack

This mix-and-match rack-scale vision includes both Xeons and Atoms with the Xeons in a different tray from the Atoms

Intel has been rethinking the rack and rather than converge all the parts of a server in a system it figures they should be disaggregated so all the compute is together, and all the shared storage is together and all the shared power and cooling is together, all the shared memory is together and everything is tied together with a high-speed optical network starting at 100 Gbps, fast enough that the elements don't have to be cheek by jowl

This mix-and-match rack-scale vision, geared to the workload, includes both Xeons and Atoms with the Xeons in a different tray from the Atoms. This way the widgetry is easier to service and upgrade. Upgrading the processors can be done separate from the other components.

It's also supposed to eliminate all unnecessary sheet metal as well as bottlenecks and cost in space and power while increasing density.

Intel's been working with a bunch of web-scale companies in China including Baidu, Tencent, Alibaba and China Telecom to realize the beginnings of this vision in what is called Project Scorpio after showing off a mechanical prototype at the complementary Facebook-led Open Compute Summit in January.

Scorpio shifts the fans and power supplies out of the individual server to a rack-lever power zone. That's phase one and involves creating a physical aggregated rack with the fan and power supplies in six zones in the rack to demonstrate reduced TCO.

Phase two means disaggregating and separating the storage from compute systems with direct attached storage, and getting higher utilization through storage virtualization. Intel says the compute and network fabric is the key technology for disaggregating the storage without impacting performance.

Intel's going to have reference designs of this architecture available for cloud service providers and large hyper-scale data centers sometime next year when the silicon photonics-based optical networking and interconnects are production-ready. That'll mean fewer cables, increased bandwidth, farther reach and power efficiency.

"Ultimately,' it says, "the industry will move to subsystem disaggregation where processing, memory and I/O will be completely separated into modular subsystems, making it possible to easily upgrade these subsystems rather than doing a complete system upgrade."

More Stories By Maureen O'Gara

Maureen O'Gara the most read technology reporter for the past 20 years, is the Cloud Computing and Virtualization News Desk editor of SYS-CON Media. She is the publisher of famous "Billygrams" and the editor-in-chief of "Client/Server News" for more than a decade. One of the most respected technology reporters in the business, Maureen can be reached by email at maureen(at)sys-con.com or paperboy(at)g2news.com, and by phone at 516 759-7025. Twitter: @MaureenOGara

Comments (0)

Share your thoughts on this story.

Add your comment
You must be signed in to add a comment. Sign-in | Register

In accordance with our Comment Policy, we encourage comments that are on topic, relevant and to-the-point. We will remove comments that include profanity, personal attacks, racial slurs, threats of violence, or other inappropriate material that violates our Terms and Conditions, and will block users who make repeated violations. We ask all readers to expect diversity of opinion and to treat one another with dignity and respect.