Weekly Wrap Podcast for October 4, 2019

Plus, Qualcomm links AI and mobile edge computing to 5G, and Docker is searching for funds

AT&T wants a powerful box for its 5G plans; Qualcomm thinks its 5G approach is unique; and Docker signals it needs money.

AT&T Lobs White Box Router Design at OCP

Qualcomm Paints Strategic Contrasts in Cloud, AI, Edge

Kubernetes Shadow Hangs Over Docker Financial Squeeze

SDxCentral Weekly Wrap Full Transcript

Today is October 4, 2019, and this is the SDxCentral Weekly Wrap where we cover the week’s top stories on next-generation IT infrastructure.

This week’s episode of the Weekly Wrap is sponsored by Silver Peak. Learn more about the Silver Peak SD-WAN solution.

AT&T submitted specifications for a distributed disaggregated chassis white box router to the Open Compute Project.

The design is targeted at the provider edge and the core routers that comprise the global IP common backbone, which is the core network that carries all IP traffic.

The specifications call for two-line cards and a fabric system.

Both of those line cards are required to support 13 400G fabric facing ports.

Those components are unique as they are to be built as standalone white boxes with their own power supplies, fans, and controllers.

They will also use external cabling instead of a backplane to interconnect the devices.

AT&T said this approach will improve horizontal scale-out as the system capacity will no longer be limited by the physical size or the backplane’s electrical conductance.

It will also simplify the cooling of critical components compared with traditional router chassis.

The specifications are based on Broadcom’s Jericho 2 system on a chip that the carrier said is ideal for the application because it has deep buffers, route scale, and port density that service providers require.

That chipset has also been optimized for 400G interfaces that are essential for the carrier as it upgrades its network to support 5G services.

Qualcomm executives think their approach to cloud computing, artificial intelligence, and mobile edge computing is unique in the market and that those technologies are going to have a big impact on 5G deployments.

Speaking at a recent event held at its headquarters in San Diego, Qualcomm CTO Jim Thompson said that the company views AI as playing a big role in its vision for what it terms the edge cloud.

He explained that Qualcomm’s focus has been on deep neural networks and deep learning for resource constrained devices like those being deployed at the edge of a network.

Thompson noted that AI is good for consuming large amounts of data, which is expected to increase as carriers deploy 5G services at the network edge.