Tuesday, June 19, 2018

ONF moves ahead with reference designs

The Open Networking Foundation (ONF) announced a series of reference designs backed by ONF’s operator leadership, which includes AT&T, China Unicom, Comcast, Google, Deutsche Telekom, Telefonica, NTT Group, and Turk Telekom.

The initial set of open source Reference Designs, which target edge and access clouds infrastructure, essentially are “blueprints” for how to put common modular components together to create platforms based on open source. They leverage the ONF's prior work with CORD and M-CORD.

Another way to think of the reference designs is as "swim lanes" for assembling open source platforms for both the cloud data center infrastructure and the underlying optical transport network.

It is also interesting to note Google's presence in ONF's operator leadership ranks. In April, Google announced plans to use P4Runtime as the foundation for its next generation of data centers and wide area network control-plane programming. P4 is a programming language was designed to be target-independent (i.e. a program written in P4 could be compiled, without modification, to run on a variety of targets, such as ASICs, FPGAs, CPUs, NPUs, and GPUs), and protocol-independent (i.e. a P4 program can describe existing standard protocols, or be used to specify innovative, new, customized forwarding behaviors).  The ONF's reference design for a Unified, Programmable & Automated Network (UPAN) will use P4 as its enabling technology. If other telcos adopt the UPAN reference design, Google will be well-positioned to offer them a wholesale service running in its cloud.

In addition, ADTRAN, Dell/EMC, Edgecore Networks and Juniper Networks have joined the ONF as supply chain partners in this reference design process.

“Over the last several years, ADTRAN has fully embraced open, disaggregated and software-defined attributes as the core principles in our Mosaic-branded access focused solutions. We are encouraged by ONF’s new strategic plan and strong operator commitment and believe ADTRAN is a perfect fit as a supply chain partner,” said ADTRAN Senior Vice President of Technology and Strategy Jay Wilson. “ADTRAN intends to contribute across multiple reference design initiatives by applying access domain expertise both in software and systems integration.”



Open Disaggregated Transport Network project gets underway

A new, operator-led Open Disaggregated Transport Network (ODTN) project is underway at the Open Networking Foundation (ONF).

The goal of ODTN is to build optical transport networks using disaggregated optical equipment, open and common standards, and open source software.

The project will deliver an open source platform for running multi-vendor optical transport networks. It will leverage the ONF’s ONOS SDN Controller, automatically and transparently discovers the disaggregated components and will control the entire transport network as a unified whole, thus enabling multi-vendor choice.

The organizers of the project say that just as the SDN movement has disaggregated the data center and operator edge networks, ODTN will bring similar benefits to the optical transport network including best-of-breed choice, elimination of vendor lock-in, cost containment and accelerated innovation.

Backers of the project include China Unicom, Comcast, NTT Communications, Telefonica and TIM.

Each of the five founding operators has committed to performing lab integration and evaluation of the platform for future transport applications. Additional support is coming from leading vendors in the optical equipment space, with NEC, NOKIA, Oplink, ZTE contributing to the software platform and building full solutions, CTTC contributing from academia, and ADVA, Ciena, Coriant, CoAdna, Infinera and Lumentum participating in lab and field trials.

Relationship to Other Projects

ODTN is the only open source solution in the optical transport space, but is leveraging other ongoing work which has focused on standardizing various interfaces and components.

ODTN will leverage and expose TAPI as its northbound interface, leveraging the work coming out of the ONF’s Open Transport Configuration and Control (OTCC) project. Likewise, OpenConfig is the base southbound model and API for communicating to optical equipment.

The OpenROADM MSA defines interoperability specifications and data models for optical devices, networks and services.  ODTN benefits from this effort and, over time, it helps the industry achieve transponder compatibility.  This will eliminate the need to deploy transponders in matched pairs, further disaggregating the solution and enabling even greater deployment flexibility.

TIP’s Open Optical & Packet Transport project is producing open DWDM architectures, models and APIs, covering transponders, open line systems, and routers. In time, the ODTN project hopes to benefit from the availability of open optical hardware coming from the TIP work.  And visa versa, the TIP project can leverage the open source work coming out of ODTN on TIP white box hardware building blocks (such as Voyager).


Mellanox supplies Infiniband for Sandia's ARM supercomputer

Mellanox Technologies will supply an InfiniBand solution to accelerate the world’s top Arm-based supercomputer to be deployed in Sandia National Laboratory in the second half of 2018.

The Astra supercomputer will include nearly 2600 nodes, and will leverage InfiniBand In-Network Computing acceleration engines. Astra is the first system in a series of the Vanguard program of advanced architecture platforms, supporting the US Department of Energy’s National Nuclear Security Administration (NNSA) missions.

“InfiniBand smart In-Network Computing acceleration engines will enable the highest performance and productivity for Astra, the first large scale Arm-based supercomputer,” said Gilad Shainer, vice president of marketing at Mellanox Technologies. “InfiniBand is the world-leading interconnect technology that supports the full range of compute and storage infrastructures, including X86, Power, Arm, GPUs, FPGAs and more. We are happy to support the Department of Energy’s efforts to explore new supercomputing platforms for their future needs.”

http://www.mellanox.com

Colt activates U.S. network

Colt has connected 13 major cities in North America, including New York, San Francisco and Chicago, to its dense Asian and European metro networks, which is made up of more than 870 data centers and 26,000 fiber-connected buildings.

Colt services available in the US include; enterprise bandwidth services up to 100Gbps, delivered over entire wavelengths and Ethernet, with private network options, and a number of wholesale services.

Colt’s On Demand bandwidth provisioning is available to businesses in Europe and Asia, with the service launching in Q4 in the US.

“Colt has been disrupting the market for more than 25 years, from our beginning as the only challenger to the local incumbents in the City of London to today, where we are a global network challenger that thinks and acts differently in a rapidly consolidating US market,” said Carl Grivner, Chief Executive Officer of Colt. “We know from our experience that business agility and the need for real-time response to customers is vital for large enterprises and financial firms. Colt is able to deliver on both counts. We’re privately held, affiliated with Fidelity Investments, and have the freedom to act extremely rapidly in a market characterized by unique, on-demand requirements.”

Facebook open sources binary optimization and layout tool

Facebook has decided to open source a binary optimization and layout tool (BOLT) that optimizes the placement of instructions in memory, thereby reducing CPU execution time by 2 percent to 15 percent.

BOLT rearranges code inside functions based on their execution profile.

Facebook said the tool was proven to be very efficient with its highly complex services which are driven by very large source code bases.

https://code.facebook.com/posts/605721433136474/accelerate-large-scale-applications-with-bolt/


DE-CIX deploys Infinera Cloud Xpress 2

DE-CIX, the world’s leading internet exchange operator, has deployed the Infinera Cloud Xpress 2 to address soaring demand for internet exchange services in rapidly growing markets including New York City, Frankfurt and Dallas.

DE-CIX serves over 1,300 carriers, internet service providers and content networks from more than 100 countries, including all leading international players in various metro markets in Europe, the Middle East, India and North America.

Infinera said DE-CIX selected its Cloud Xpress 2 for the combination of capacity and simplicity in a compact, easy-to-install optical networking platform.

Infinera's Cloud Xpress 2 supports simple point-and-click provisioning. It delivers 1.2 terabits per second of capacity up to 130 kilometers without requiring any external equipment.  Infinera Instant Bandwidth allows DE-CIX to activate and pay for additional capacity in minutes, reducing the initial cost of deployment while remaining responsive to customer capacity demands.

The solution was implemented by network specialist Axians Networks & Solutions, which is a longtime partner of Infinera and has been involved in other projects for DE-CIX.

"DE-CIX is experiencing tremendous and unpredictable customer demand for interconnection services in major markets across the globe," said Daniel Melzer, CTO, DE-CIX. "Our customers request instant activation of high-capacity internet exchange services, and Cloud Xpress 2 with Instant Bandwidth enables us to deliver that to our customers. Cloud Xpress 2 is easy to install and is performing better than advertised for our data center interconnection requirements."

"With internet traffic booming, driven by growth in cloud-based services, neutral interconnection services from DE-CIX are the glue connecting the cloud ecosystem," said Nick Walden, Infinera Senior Vice President, EMEA. "We are delighted to work with DE-CIX to rapidly meet the needs of their customers.”

ADVA's Oscilloquartz delivers enhanced PRTC for 5G network timing

Oscilloquartz, which is a division of ADVA, launched an enhanced primary reference time clock (ePRTC) system for 5G network timing.

The Oscilloquartz solution, which is able to provide timing accuracy and stability even when the GNSS signal is lost, features the OSA 3230B ePRC atomic cesium clock connected to an Oscilloquartz clock combiner and grandmaster.

The company said the technology provides the ideal timing source for mission-critical transport systems, such as utility networks, government infrastructure and radio access networks, and provides the strict synchronization needed for LTE-A and 5G applications.

“With our ePRTC system, we’re taking reliability and accuracy to the next level. This solution meets the requirements of next-generation mobile networks, offers the precise synchronization needed by many of today’s industries and removes vulnerability to GNSS outages,” said Gil Biran, general manager, Oscilloquartz. “Combining our advanced multi-constellation GNSS receiver with our atomic cesium clock technology creates an outstandingly accurate source of time with guaranteed holdover performance. This provides vital protection against loss of satellite signal due to jamming, which can be a major problem."

HPE to invest $4 billion in Intelligent Edge technologies

Hewlett Packard Enterprise plans to invest $4 billion in Intelligent Edge technologies and services over the next four years.

The plan calls for investments in research and development to advance and innovate new products, services and consumption models across a number of technology domains such as security, AI and machine learning, automation and edge computing. HPE said it intends to build out its portfolio of transformational advisory and professional services for the edge.  HPE will also continue to advance Memory-Driven Computing. The company vowed support for open standards and open source technologies.

“Data is the new intellectual property, and companies that can distill intelligence from their data —whether in a smart hospital or an autonomous car—will be the ones to lead,” said Antonio Neri, president and chief executive officer, HPE. “HPE has been at the forefront of developing technologies and services for the Intelligent Edge, and with this investment, we are accelerating our ability to drive this growing category for the future. The next evolution in enterprise technology will be in edge-to-cloud architecture. Enterprises will require millions of distributed clouds that enable real-time insights and personalized experiences exactly where the action is happening.”

Cisco to acquire July Systems for location services platform

Cisco agreed to acquire July Systems, a start-up offering a cloud-based mobile experience and location services platform. Financial terms were not disclosed.

July Systems, which based in Burlingame, California with offices in Bangalore, India, has worked for several years as an OEM for Cisco Connected Mobile Experience (CMX). Cisco plans to add July Systems’ platform and business context capabilities to provide a unified solution on which partners and customers can build and deliver indoor location services for industries as diverse as healthcare, government, logistics, manufacturing, sports arenas, hotels, education and retail.

Cisco said the acquisitions supports its journey to intent-based networking.

The July Systems team will join Cisco’s Enterprise Networking Group led by Scott Harrell, senior vice president and general manager.

Orange conducts Cloud RAN trial with Nokia

Orange conducted a trial of Cloud RAN technology in its live network in Poland in conjunction with Nokia.

The trial, which used Orange's own cloud infrastructure, took place from March to the beginning of May in Poland, with radio sites in the city of Chelm and the virtualized part of the baseband running in a data center in Lublin, around 70 km away.. The two companies trialed Cloud RAN technology to prepare for the eventual introduction of a distributed cloud architecture for 5G by Orange.

Nokia's architecture splits baseband processing functionality across the radio cell sites and data center using its AirScale Cloud Base Station. Time-critical functions are performed at the cell site and connected via Ethernet fronthaul - allowing the operator to use its existing transport network - while centralized software hosted at the data center cost-efficiently performs non real-time functions.

About the trial

  • Nokia AirScale Cloud Base Station, a virtual base station for 4G and 5G, running 4G technology via Ethernet fronthaul
  • Orange's NGPop cloud infrastructure
  • Nokia AirFrame Data Center platform designed to meet stringent radio access capacity, performance and latency requirements
  • Nokia Cloud Infrastructure for Radio/Real Time
  • Nokia NetAct monitoring Cloud RAN and classical RAN in the commercial network

Piotr Jaworski, CTO of Orange Poland, said: "During this trial, Orange Poland has successfully trialed virtualized RAN architectures in collaboration with Nokia and investigated the impact on network operations. The quality and performance observed during the trial give confidence on the RAN virtualization and is a critical step for us towards 5G."

HPE debuts GreenLake Hybrid Cloud

HPE introduced its GreenLake Hybrid Cloud for managing its enterprise customers' private infrastructure while optimizing integration with AWS, Microsoft Azure and Azure Stack.

Unlike a traditional managed service, HPE said its GreenLake will provide an automated, cloud-native model that eliminates the need for organizations to hire or train new staff to oversee and manage cloud implementations.

"Enterprises need a fast, simple way to optimize their public and on-premises clouds to enable applications and data under a common operating model,” said Ana Pinczuk, SVP and GM of HPE Pointnext. “HPE GreenLake Hybrid Cloud is a breakthrough solution that removes the complexity of operating multiple clouds and enables customers to maintain control, reduce costs and accelerate time to value."

https://www.hpe.com/us/en/services/it-consumption.html

Toshiba launches lower cost SAS SSDs for servers

Toshiba introduced a new category of SAS SSDs to replace SATA SSDs in servers.

The RM5 12Gbit/s value SAS (vSAS) series feature Toshiba's BiCS FLASH TLC (3-bit-per-cell) 3D flash memory and will initially be available in capacities up to 7.68TB in a 2.5” form factor.

Toshiba said the new drives offer capacity, performance, reliability, manageability and data security advantages – at a price that obsoletes SATA SSDs.

See also