Wednesday, March 22, 2017

Facebook shows its progress with Open Compute Project

The latest instalment of the annual Open Compute Project (OCP) Summit, which was held March 8-9 in Silicon Valley, brought new open source designs for next-generation data centres. It is six years since Facebook launched OCP and it has grown into quite an institution. Membership in the group has doubled over the past year to 195 companies and it is clear that OCP is having an impact in adjacent sectors such as enterprise storage and telecom infrastructure gear.

The OCP was never intended to be a traditional standards organisation, serving more as a public forum in which Facebook, Microsoft and potentially other big buyers of data centre equipment can share their engineering designs with the industry. The hyper-scale cloud market, which also includes Amazon Web Services, Google, Alibaba and potentially others such as IBM and Tencent, are where the growth is at. IDC, in its Worldwide Quarterly Cloud IT Infrastructure Tracker, estimates total spending on IT infrastructure products (server, enterprise storage and Ethernet switches) for deployment in cloud environments will increase by 18% in 2017 to reach $44.2 billion. Of this, IDC estimates that 61% of spending will be by public cloud data centres, while off-premises private cloud environments constitute 15% of spending.

It is clear from previous disclosures that all Facebook data centres have adopted the OCP architecture, including its primary facilities in Prineville (Oregon), Forest City (North Carolina), Altoona (Iowa) and LuleƄ (Sweden). Meanwhile, the newest Facebook data centres, under construction in Fort Worth (Texas) and Clonee, Ireland are pushing OCP boundaries even further in terms of energy efficiency.

Facebook's ambitions famously extend to connecting all people on the planet and it has already passed the billion monthly user milestone for both its mobile and web platforms. The latest metrics indicate that Facebook is delivering 100 million hours of video content every day to its users; 95+ million photos and videos are shared on Instagram on a daily basis; and 400 million people now use Messenger for voice and video chat on a routine basis.

At this year's OCP Summit, Facebook is rolling out refreshed designs for all of its 'vanity-free' servers, each optimised for a particular workload type, and Facebook engineers can choose to run their applications on any of the supported server types. Highlights of the new designs include:

·         Bryce Canyon, a very high-density storage server for photos and videos that features a 20% higher hard disk drive density and a 4x increase in compute capability over its predecessor, Honey Badger.

·         Yosemite v2, a compute server that features 'hot' service, meaning servers do not need to be powered down when the sled is pulled out of the chassis in order for components to be serviced.

·         Tioga Pass, a compute server with dual-socket motherboards and more IO bandwidth (i.e. more bandwidth to flash, network cards and GPUs) than its predecessor, Leopard, enabling larger memory configurations and faster compute time.

·         Big Basin, a server designed for artificial intelligence (AI) and machine learning, optimised for image processing and training neural networks. Compared to its predecessor, Big Basin can train machine learning models that are 30% larger due to greater arithmetical throughput and by implementing more memory (12 to 16 Gbytes).

Facebook currently has web server capacity to deliver 7.5 quadrillion instructions per second and its 10-year roadmap for data centre infrastructure, also highlighted at the OCP Summit, predicts that AI and machine learning will be applied to a wide range of applications hosted on the Facebook platform. Photos and videos uploaded to any of the Facebook services will routinely go through machine-based image recognition and to handle this load Facebook is pursuing additional OCP designs that bring fast storage capabilities closer to its compute resources. It will leverage silicon photonics to provide fast connectivity between resources inside its hyper-scale data centres and new open source models designed to speed innovation in both hardware and software.

Luxtera unveils Duplex 100G CWDM2 QSFP with PAM4

Luxtera, a leading supplier of silicon photonics solutions, announced the sampling of its 100G-CWDM2 QSFP optical transceiver module based on its latest silicon photonics chipset and targeting applications in cloud, enterprise and high performance computing,

Luxtera's new LUX45202 optical transceiver is a 100 Gbit/s duplex solution developed for 100 Gbit/s deployments for cloud, enterprise and high performance computing end users, and is designed to offer high performance and reliability at a low price point. The solution implements PAM4 modulation, enabling the device to deliver using two lasers the performance that current solutions require four lasers to achieve.

The new LUX45202 transceiver module offers key features including:

1. 100 Gbit optical transceiver in QSFP28-compliant form-factor.

2. Multi-rate capability, with support for 10.3125, 20.625 or 25.78125 Gbit/s per electrical channel.

3. Support for CWDM2 and reach of over 2 km.

4. Power consumption of less than 3.5 W.

Luxtera's LUX45202 100 Gbit/s duplex CWDM2 optical transceiver is currently sampling to select customers, with volume production scheduled for the second half of 2017.

Recently, Luxtera announced a partnership with TSMC for next generation silicon photonics manufacturing. Under the agreement, Luxtera and TSMC are developing a new silicon photonics technology platform designed to double optical link performance while enabling four times the raw data rate of existing silicon photonics solutions. The technology will provide SoC integration of optical interconnect with the CMOS logic for applications including network switch, storage and compute solutions.

Cavium and Elenion offer silicon photonics-based reference architecture

Cavium, a provider of semiconductor products designed to enable secure, intelligent packet processing, and Elenion Technologies, a developer of silicon photonics technology, announced an end-to-end reference architecture for use in enterprise and cloud data centres.

Integrating Cavium's QLogic network interface cards (NICs) and XPliant-based switches, with Elenion's silicon photonics on-board optics, the direct fibre-to-the-server and parallel multi-channel on-board solution for switches is designed to provide increased bandwidth to the server, aggregation at 400 Gbit/s and beyond, lower power consumption and offer a path to achieving the target of less than $1 per Gbit.

For the joint solution, Cavium is contributing its QLogic high performance, multi-protocol Ethernet adapters, which support 10/25/40/100 Gbit/s speeds, and the XPliant Ethernet switch family, providing a high throughput, programmable data centre switching solution.

Cavium noted that platforms based on the XPliant Ethernet switches are able to leverage its programmatic control of table resources and pipeline logic to address the specific needs of the network architecture, while also providing a high degree of packet visibility and telemetry.

Cavium and Elenion are displaying the reference design for fibre-to-the-server solutions during OFC 2017 in Los Angeles. A demonstration based on Elenion's next-generation photonics platform and packaging technology, combined with the Cavium solutions, showcases advanced on-board transceivers with light sources integrated onto the silicon photonics to enable a scalable, cost disruptive on-board solutions for NIC and switch applications within the data centre.

Recently, Coriant introduced a short reach CFP2-ACO pluggable module for its Groove G30 Network Disaggregation Platform (NDP) that included silicon photonics technology from Elenion, The Groove short reach CFP2-ACO pluggable is designed to enable cost-optimised, power-efficient 200 Gbit/s connectivity for carrier transport and DCI applications.

EXFO launches FTB-4 Pro platform for spectral and transport

EXFO has announced the launch of the FTB-4 Pro platform for 100 Gbit/s networks, which it claims is the most compact platform for testing high speed and optical networks, expanding the company's test orchestration portfolio designed to enable efficient and scalable network testing in support of field operations, network performance assurance and service delivery.

EXFO's new 4 slot modular FTB-4 Pro platform combines support for 100 Gbit/s commissioning, turn-up and troubleshooting. The solution includes the FTBx-88200NGE 100 Gbit/s multi-service tester featuring its iOptics transceiver validation software, and the FTB-5240S-P optical spectrum analyser. The solution is designed to eliminate the need for additional platforms or swap modules by enabling transport and spectral testing using a single platform.

The FTB-4 Pro platform supports a range of modules for field testing, data centre interconnect, submarine testing and lab applications to provide increased flexibility and support testing across all phases of the service delivery chain, from development to deployment, maintenance and troubleshooting.

Additional module combinations provided by the FTB-4 Pro platform also include iOLM/OTDRs, OLTS for fibre characterisation, dispersion solutions and other transport modules, all of which offer compatibility with EXFO's fibre inspection probes.

EXFO noted that the FTB-4 Pro platform features a 10-inch high-resolution widescreen display and a QUAD core processor and operates on the latest Microsoft Windows 10 OS.

Menara samples C-band tunable coherent CFP-DCO transceivers

Menara Networks, an IPG Photonics business and vertically integrated provider of IP/Ethernet DWDM transport solutions, has announced immediate sampling of its coherent CFP-DCO transceivers to select customers.

Based on an advanced 16 nm coherent processor, Menara's latest digital CFP-DCO DWDM transceiver is designed to support DWDM coherent transmission for 100 and 200 Gbit/s per wavelength in a CFP MSA form factor. The device performs functionality including adaptive advanced modulation, digital signal processing for linear and non-linear fibre impairments compensation, G.709-compliant OTN framing and a range of hard and soft decision-based FEC encoding/decoding.

Compatible with 100 Gigabit Ethernet, CAUI, OTL4.10 and OTL 4.4 interfaces, Menara's new coherent CFP-DCO is full C-band tunable and offers support for ITU-T G.694.1 12.5 GHz flexible grid. The solution is designed to support ZR, data centre interconnection (DCI), metro and long haul network applications.

At OFC 2017, Menara is showcasing its new coherent CFP-DCO transceivers on a 250 km repeater-less fibre link featuring IPG Photonics' compact hybrid EDFA/Raman Optical Services Transport Platform (OSTP). The company will also demonstrate its transceiver solution integrating C-band tunability, OTN framing and FEC in a SFP+ transceiver MSA form factor.

In March 2016, Menara announced the availability of a MSA CFP-2 100 Gbit/s adapter module designed to accommodate QSFP28 transceivers covering all 100 Gbit/s MSAs and reaches. Menara claimed this was the first solution able to adapt any MSA QSFP28 100 Gbit/s module into a MSA-compliant CFP2 100 Gbit/s module.

FatPipe enhances SD-WAN with L2/3 load balancing

FatPipe Networks, the developer of software-defined networks for wide area connectivity and hybrid WANs, has announced its Next Gen SD-WAN platform, version 9, which is designed to simplify the management of a WAN and provide load balancing for Layer 2 and 3 networks, thereby eliminating the need for routers.

FatPipe's Next Gen Version 9 SD-WAN platform is designed to enable customers to fully integrate its SDN at the data centre with the SD-WAN at the branch level, providing a scalable SDN + SD-WAN solution that is claimed to offer superior functionality to current SD-WAN products. In addition, virtual versions are available for Amazon and Azure platforms, plus VMWare and other popular platforms.

FatPipe's latest Next Gen SD-WAN platform is designed to deliver feature including:

1. Layer 2 and 3 support, enabling routed and switched topologies for global SDN deployments that bridge the LAN and WAN, while also providing low-latency performance and enabling lower-cost deployments.

2. New configuration design with a web-based interface that scales for any screen and offers a unified view across the network and a simplified configuration view to help bring appliances online.

3. New auto-configuration functionality designed to reduce the complexity of configuring multiple appliances, with MPSec, VPN and policy routing rules propagated to remote/branch devices automatically from a central location, plus improved support for legacy routing protocols via BGP and OSPF support.

4. Application visibility and performance allow application flows to be managed across the WAN, with enhanced flexibility in application identification and control over flow direction and a holistic view across the network providing granular application visibility.

FatPipe's 'branch in a box' SD-WAN product integrates all branch requirements, including multi-line load balancing, firewall, VPN, QoS and local Smart DNS. The company noted that with Smart DNS hosted at the local branch, if the cloud-based management product, FatPipe Orchestrator, is inaccessible, the local appliances can continue to perform and transmit data.

Tuesday, March 21, 2017

Ciena Opens WaveLogic to Lumentum, NeoPhotonics and Oclaro

Ciena announced it is extending the market reach of its WaveLogic coherent modem technology by making it available to a number of industry participants, with optical component companies Lumentum, NeoPhotonics and Oclaro to leverage WaveLogic technology to develop coherent optical modules for a range of applications.

Additionally, each of the companies will join an ecosystem that will focus on standards-based product development, designed to provide customer choice in the emerging 400 Gbit/s pluggable market, which is being driven by increasing data centre interconnect (DCI) requirements.

Ciena stated that while retaining its intellectual property rights, through the initiative it will provide its WaveLogic Ai chipset to Lumentum, NeoPhotonics and Oclaro. The partners will each be responsible for the manufacture, sale and distribution of Ciena-designed 5 x 7 coherent modules based on a combination of the WaveLogic technology and electro-optics components.

In addition, under the agreement the partner companies will contribute to the ongoing development of 400 Gbit/s pluggable technologies and the establishing of specifications for adoption by relevant industry forums and standards-setting bodies. Ciena intends to invest in further developing coherent DSP technology, which its partners will then be able to utilise in the development of 400 Gbit/s pluggable modem solutions for hardware and systems vendors.

Lumentum, NeoPhotonics and Oclaro will also support the creation of a roadmap for future generations of coherent optical modules based on Ciena's WaveLogic technology.

Separately, Oclaro announced it has been selected by Ciena as a partner for the development of 400 Gbit/s coherent networks leveraging Ciena's 56 Gbaud WaveLogic Ai coherent DSP chipset. The technology will be used to develop programmable modules supporting up to 400 Gbit/s rates.

Oclaro stated that it expects to sample its first 400 Gbit/s flex coherent transponders based on the WaveLogic Ai DSP chipset by the end of the year. The device will feature a single wavelength transmitter in a 5 x 7 inch form factor designed to support flexible bandwidth for 400, 300, 200 and 100 Gbit/s applications.

Coriant Unveils Aware Toolkit for Real-time Optical Performance, Control Capability

Coriant has announced the introduction of Coriant Aware Technology, an new software-enabled toolkit designed to enhance the value of dynamic optical networks via sophisticated real-time performance awareness and control functionality.

In a live demonstration at OFC 2017, Coriant will showcase how the Aware technology can be used to enable faster and simpler new wavelength planning and provisioning, increased reach and capacity and improved service availability.

Coriant noted that existing networks suffer from limitations that can increase costs, increase wavelength activation times, limit reach and reduce efficiency, with these limitations in part due to planning tools that do not have real-time data from the optical network, as well as coherent receivers that are unable to measure OSNR and NMS, SDN and ASON/GMPLS implementations that lack the optical models of offline planning tools. In addition, the adoption of SDN and flexi-rate interfaces will further emphasise these limitations.

The new Coriant Aware technology is designed to correlate, analyse and act upon real-time data by applying concepts already used for the Internet of Things (IoT) and help to address these limitations, as well as help operators reduce capex and opex.

Coriant noted that residual margin in a network is affected both by OSNR and linear and non-linear impairments, so that accurately assessing residual margin means determining the impact of all three factors. For example, while current coherent receivers may be able to measure linear effects such as CD, PMD and PDL, they cannot measure the impact on signal quality, nor accurately measure OSNR or the effect of nonlinear penalties.

To address these issues, Aware technology features two components: the Optical Performance Engine (OPE) and Margin Processing Engine (MPE).

Optical Performance Engine

OPE is designed to combine the accuracy of an offline planning tool with the real-time speed needed for wavelength activation, Layer 0 restoration, and other optical use cases, with real-time performance enabled by simplifying the mathematical operations.

The OPE responds to path validation requests from the path computation function within the NMS, SDN or ASON/GMPLS control plane. With optical models that incorporate both linear and nonlinear effects, the OPE generates valid options for each requested path, including modulation types, baud rates, FECs, frequencies and power levels to allow the optimum options, including flexi-rate interface settings and super-channels, to be selected.

Margin Processing Engine

MPE is designed to enable operators to accurately estimate residual margin by collecting performance monitoring data from across the network, including coherent receivers and per-channel power monitoring capabilities of the network elements. It then processes this data in real-time to distinguish OSNR from signal degradation due to linear and nonlinear impairments, providing accurate, real-time residual margin values for each channel.

Additionally, MPE is able to provide this without the need for next-generation DSPs, allowing use with installed coherent interfaces.

Key capabilities enabled by Coriant Aware technology include:

1. Increased transmission reach and capacity by reducing or eliminating margin stacking and optimising power levels.

2. Faster wavelength service activation through reducing the provisioning steps required and the boundaries between planning and management systems.

3. Higher availability via support for SDN or ASON/GMPLS restoration, enabling a reduction in margin requirements for temporary restoration paths and pro-active downtime prevention.

4. Allows operators to use margin as extra capacity that can be provided to customers.

Coriant Aware technology will be introduced in phases, with initial OPE functionality due to be available for trials beginning in the second quarter of 2017.

Oclaro Unveils I-Temp Transceivers for 5G fronthaul, Samples 400G

Oclaro announced the introduction of two new industrial temperature (I-temp) transceivers designed to drive the market transition to 5G wireless fronthaul systems and sampling of its 400 Gbit/s CFP8 transceiver for core routers/transport applications.

I-temp transceivers

Oclaro's new I-temp transceivers are designed to provide customers with the flexibility to choose either a 10 Gbit/s TSFP+ I-temp that can fit a WDM fronthaul system into a next generation 5G wireless network, or a 25 Gbit/s SFP28 LR I-temp that significantly increases data capacity and offers industry-leading low power consumption.

The TSFP+ is Oclaro's latest product to include I-temp and follows the company's 10 Gbit/s TSFP+ C-temp and E-temp devices, which are already shipping. The TSFP+ is vertically integrated based on Oclaro's in-house ILMZ chip and custom-designed TOSA and TSFP+ technology. Oclaro has enabled 85 degrees C operation by optimising the operation temperature of the ILMZ chip and -40 degrees C operation by developing TEC to lower power consumption, as well as improving power dissipation of key components.

Oclaro expects to begin sampling the 10 Gbit/s TSFP+ I-temp in April, with volume production due to start in the third quarter of 2017.

400 Gbit/s CFP8 transceiver

Oclaro's 400 Gbit/s CFP8 transceiver targets core routers/transport applications and leverages its EML laser and receiver technology to deliver higher data rates between high-end routers and optical transport systems. At OFC Oclaro will demonstrate the CFP8 operating with 8 lasers at 50 Gbit/s PAM4.

With a 40 x 102 x 9.5 mm form factor, the 400 Gbit/s CFP8 transceiver provides a high density, high-throughput solution. The CFP8 leverages technology from Oclaro's earlier CFP, CFP2 and CFP4 designs, including its cooled 1310 nm 28 Gbit/s EA-DFB lasers, integrated TOSA and ROSA and 28 Gbit/s 4 channel integrated PIN-PD array.

The CFP8 transceiver for reach of 2 km and 10 km is compliant with 400GBASE-LR8 optical and 400GAUI-16 electrical interface specifications, offers EML laser technology offering good margin over IEEE 400 Gigabit Ethernet optical specifications and receiver bandwidth capability to allow interfacing with different lasers.

Oclaro's CFP8 product is currently sampling, with volume production scheduled for the second half of 2017.

Inphi Launches ColorZ 100G PAM4 platform, unveils 16 nm CMOS PAM4

Inphi, a provider of high-speed interconnect solutions, announced the commercial availability and production ramp of ColorZ, which it claims is the first silicon photonics 100 Gbit/s PAM4 platform for 80 km DWDM data centre interconnect (DCI) in QSFP28 form factor, and introduced the Polaris 16 nm CMOS PAM4 solution.


First unveiled at OFC 2016, Inphi's ColorZ utilises advanced PAM4 signalling and is able to support up to 4 Tbit/s of bandwidth over a single fibre, allowing multiple data centres located within 80 km of each other to be connected and behave as a single data centre facility.

The ColorZ DWDM platform offers key features including: support for 100 Gigabit Ethernet; extended reach up to 80 km; 4.5 W power dissipation; compatibility with ITU wavelength grid; InphiNity core DSP engine with a mixed-mode DSP architecture for low power applications requiring adaptability and configurability; and multiple programmable FEC options, with varying levels of pre-FEC BER performance.

At OFC 2017, Inphi is carrying out live interoperability demonstrations of its ColorZ QSFP28 solution, including driving 3.6 Tbit/s of bi-directional east-west traffic over 80 km of fibre between two Arista 7500R platforms, and 100 Gbit/s line cards connected via AVDA direct detect Open Line System (OLS) with SmartAmp technology.

The demonstration is designed to showcase the simplicity and plug-and-play capabilities of the ColorZ technology, and to demonstrate its ability to offer a cost-effective solution for edge data centre connectivity.


Inphi's new Polaris offering is claimed to be the first 16 nm CMOS PAM4 solution for next-generation cloud deployments and includes the company's integrated, low power PAM4 DSP IC and low power linear driver and TIA for data centre connectivity.

Inphi's 16 nm PAM4 DSP PHY ICs provide a full bi-directional interface with host ASICs that have 28 GBaud PAM4 and NRZ electrical interfaces and support bridging to 28 GBaud PAM4 optics. The products also support retiming and gearbox functionality, with packaging designed for the following optics modules:

1. Polaris-400G, 8 x 56 Gbit/s PAM4, 16 x 28 Gbit/s NRZ gearbox for CFP8.

2. Polaris-400G (NG), 8 x 56 Gbit/s, 8 x 56 Gbit/s PAM4 retiming for OSFP/QSFP-DD.

3. Polaris-200G, 4 x 56 Gbit/s, 4 x 56 Gbit/s PAM4 retiming for QSFP56.

4. Polaris-100G, 2 x 56 Gbit/s, 2 x 56 Gbit/s PAM4 retiming for QSFP28.

5. Polaris-50G, 1 x 56 Gbit/s, 1 x 56 Gbit/s PAM4 retiming for SFP56.

The Polaris products also feature the InphiNity DSP engine with mixed-mode DSP architecture, a range of self-test and loopback modes for diagnostic monitoring of channel and system parameters, eye-scan, samplers and monitors on receiver interfaces, and API routines for data and error analytics.

Inphi noted that its IN2834 28 Gbaud low power quad linear driver: and IN2864 28 Gbaud low power quad linear TIA for PAM4 optical modules are currently sampling to lead customers.

Source Photonics Demos XGS-PON N2 OLT module, 53 Gbaud PAM4

Source Photonics announced that at OFC 2017 it is demonstrating its recently introduced XGS-PON N2 OLT optical module, and that with MACOM Technology Solutions (MACOM) it is showcasing a 53 Gbaud PAM4 solution delivering a 6 dB loss budget using a compact TOSA and ROSA capable of supporting single lambda 100 Gbit/s in a QSFP28 form factor.

Source Photonics' XGS-PON optical transceiver module is designed to offer a universal upgrade path from existing GPON networks to next generation XGS-PON networks supporting multi-gigabit services. The company is demonstrating the transmission with the XGSPON N2 OLT XFP optical module over a 20 km optical link.

The XGS-PON N2 OLT offered in the XFP form factor meets the N2 class maximum optical link budget set in the XGS-PON standard G.9807.1, providing 9.953 Gbit/s downstream and upstream data rates and supporting a 1:64 split ratio over 20 km links.

The company noted that its optical and electrical design capabilities enable the integration of a 1577 nm EML transmitter and burst mode 1270 nm APD receiver with micro-optic WDM filters to provision a single-fibre, dual-wavelength upgrade and overlay pathway for migration of existing class B+ and C+ GPON networks to XGPON and XGS-PON. Source Photonics will demonstrate downstream optical power of the OLT at over 5 dBm with uplink sensitivity of -30 dBm at 1 x 10-3.

The XGS-PON N2 OLT is due to begin sampling in March 2017 and available for production in the third quarter of the year.

In addition, MACOM and Source Photonics are demonstrating a 53 Gbaud PAM4 solution. The companies are part of the ecosystem for 53 Gbaud PAM4 components that is aiming to meet demand for lower cost, higher density solutions for cloud data centre applications leveraging single wavelength 100 Gbit/s PAM4 technology.

The demonstration is designed to show that 53 Gbaud PAM4 PHYs can enable a production-grade EML laser to replace four 25 Gbit/s NRZ lasers; thereby improving the scalability of 100 Gbit/s transceivers, as well as providing the basis for duplex single mode fibre for 400 Gbit/s small form factor modules.

The demonstration comprises MACOM's 16 nm, 56 Gbaud PAM4 PHY IC on an evaluation board driving an optical loop back through a Source Photonics TOSA and ROSA. The small form factor TOSA package developed by Source Photonics uses its internal 53 Gbaud EML laser, which can deliver a more than 3 dB improvement over the proposed 100GBASE-DR OMA specification.

The joint 53 Gbaud PAM4 demonstration utilising a compact TOSA and ROSA combination, is designed to operate below the KP4 FEC requirement with a 6dB loss budget. Source Photonics and MACOM believe that the demonstration marks a key step toward realising a commercial single lambda solution able to replace many four channel solutions.

Source Photonics also announced that in cooperation with Semtech it is demonstrating a 53 GBaud PAM4 compact ROSA capable of supporting a single lambda 100 Gbit/s QSFP28. The demonstration comprises an optical loop back from a Source Photonics' 53 GBaud TOSA using a production-grade EML laser and 53 GBaud ROSA based on Semtech's GN1089 TIA. The combined TOSA and ROSA link can support a loss budget of 5 dB and better than 1 x 10-5 BER over the link.

InnoLight Demos 400G OSFP

InnoLight Technology based in Suzhou, China, a supplier of next generation cloud computing optical transceivers, has announced volume shipments of a suite of QSFP28 products and the availability of what it claims is the first 400 Gbit/s OSFP at OFC 2017 in Los Angeles.

InnoLight noted that 100 Gbit/s QSFP28 has become the de-facto standard for optical interconnects for cloud networks, as on a cost-per-gigabit basis 100 Gbit/s QSFP28 can be significantly more cost effective compared with 10 Gbit/s SFP+ and 40 Gbit/s QSFP+. As a result, the industry is rapidly deploying QSFP28 connections.

To support this transition, InnoLight is introducing a complete portfolio of QSFP28 optics solutions, as follows: SR4; eCWDM4; eSR4; CLR4; AOC; LR4 (Ethernet); QSFP28 to 4 x 25 Gbit/s SFP28 AOC; LR4 (dual rates); PSM4; LR4 (receive only); CWDM4; and ER4-Lite.

InnoLight stated that it is currently offering seventeen QSFP28 parts in mass production. The company noted that it is developing 400 Gbit/s solutions based on both the OSFP and QSFP-DD form factors to enable the optimal price-performance.

In addition, to support the 400 Gbit/s expansion, InnoLight will conduct a live demonstration of 400 Gbit/s OSFP optics during OFC.

PacketLight establishes channel partner program for metro and DCI markets

PacketLight Networks, a provider of optical networking solutions, announced the launch of its Channel Partner Program (CPP) for enterprise VARs and resellers, designed to expand its footprint of bundled solutions for metro and DCI optical networks.

The new PacketLight program offers members the product training, customisation, pre- and post-sales support and incentivised pricing.

PacketLight's CPP features five tiers and includes: training, both technical and sales, designed to help partners understand customers' requirements for WDM solutions; and custom design blueprints, with the sales support team providing network designs, assisting with technical presentations and demonstrations for customers. It also provides partners with technical support, 24 x 7.

PacketLight's products are designed to provide optical transport for high speed, secure data centre interconnect (DCI) and metro networks. Its technology is designed to increase spectral efficiency, optimise wavelength resources and enable higher fibre and metro network utilisation.

In February, PacketLight announced a partnership with Avelacom to build a high speed network from London to Moscow using its optical transport solutions. The PacketLight solution was designed to reduce network latency and enable the company to meet the demands of capital markets for low latency data transport across the continent.

Avelacom specifically selected PacketLight's PL-1000GT muxponder/transponder solution, and PL-1000IL platform, to establish a 100 Gbit/s low latency DWDM long-haul network spanning 1,100 km, with full add/drop capability at all major sites.

In November 2016, PacketLight launched the PL-2000M transport solutions for DCI and metro networks. The new muxponder/transponder is claimed to enable 30% lower cost and 2x increase in the spectral efficiency for improved network utilisation. The product supports carrier-grade coherent 200 Gbit/s tunable uplink, capable of serving applications and protocols including data, storage, OTN and TDM.

Monday, March 20, 2017

Infinera's Instant Network Delivers Software Defined Capacity for Optical Transport

Infinera introduced Instant Network capabilities for delivering software defined capacity (SDC) for cloud scale networks.

The new capabilities extend Infinera Instant Bandwidth, which provides software activation of service-ready optical capacity in a few hours on networks powered by its DTN-X platform and equipped with its unique photonic integrated circuits (PICs). Infinera introduced Instant Bandwidth in 2012 and more than 70 Service Providers are now using it. In 2015, Infinera introduced Time-based Instant Bandwidth in 2015, enabling software activation of bandwidth licenses for a limited duration.

Instant Network adds Bandwidth License Pools, Moveable Licenses and Automated Capacity Engineering (ACE). It also extends SDC to new Infinera platforms, which support flexible grid and sliceable 2.4 terabit super-channels powered by the Infinera Infinite Capacity Engine.


  • Bandwidth License Pools: The new Bandwidth License Pool functionality enables service providers to activate capacity at the same time as an invoice is issued for that capacity, reducing capital expenditures for idle capacity. This new capability enables capacity to be deployed in minutes, while previously, service providers would purchase a license before the capacity was software-activated, which resulted in deployment taking a few hours.
  • Moveable Licenses: New Movable Licenses enable service providers to use software to move bandwidth licenses across the network as traffic conditions change or fiber cuts occur. Previously, licenses were fixed to a specific line module or platform. Moveable licenses reduce capital expenses by reducing idle capacity provisioned specifically for network resiliency and increase agility when responding to customer demands for new services.
  • Automated Capacity Engineering (ACE): The new ACE application takes previously manual offline route and capacity planning processes and implements those algorithms in a microservices-based path computation element (PCE). ACE understands optical impairments and computes optimal Layer 0 routes between nodes across multiple paths, including automatic routing and wavelength assignment with multiple path constraints such as traffic engineering cost, distance and latency.

Infinera Instant Network capabilities are planned across the Infinera DNA software and the Xceed Software Suite. Instant Network Bandwidth License Pool and Moveable Licenses are available now. ACE is planned for 2018.

Infinera said its Instant Network  enables service providers to automate optical capacity engineering and scale optical capacity in minutes by using Infinera’s Xceed and Digital Node Administrator (DNA) software.  It also sets a foundation for cognitive networking, which includes advanced analytics, machine learning from streams of network telemetry data, autonomous operation of routine tasks, predictive analysis of network problems before they occur and proactive recommendations for network optimization to further reduce operational expense and improve service reliability.

“Infinera is committed to providing our customers with innovative solutions to help them win in the markets they serve,” said Dr. Dave Welch, Infinera co-founder and president. “Software defined capacity is a requirement to realize the benefits envisioned by SDN and network functions virtualization. Infinera Instant Network extends our success with Instant Bandwidth, leading the market for software defined capacity while lighting the way to cognitive networking.”

Fujitsu Enhances and Expands 1FINITY Family for DCI

Fujitsu Network Communications announced enhancements to its 1FINITY T100, including a hyper-dense transponder packaged together with the new 1FINITY T400, a 1 Tbps layer 1 aggregator. By

The company said that by decoupling the transponder and the aggregator, both of which are based on open software and open optics, it can offer customers greater architectural flexibility and industry-leading 10GbE to 100G aggregation density for data center interconnect applications. When the T400 and T100 are deployed together, the unique design of the “decoupled” aggregator and transponder provides ten times better 10GbE to 100G aggregation density than existing solutions.

Some highlights:

  • The 1FINITY T100 Transport blade, which is designed for 100G and 200G coherent optical transport between data centers in point-to-point or ring topologies, can be deployed independently, in conjunction with other 1FINITY blades such as the L100 Lambda series to provide ROADM capability, or over a third-party FOADM. Numerous feature enhancements have been added to the T100 platform. Zero touch provisioning automates configuration and provisioning, significantly speeding installation and reducing operating expenditures. Streaming telemetry technology provides an improved view of network performance without taxing software management resources to such an extent that other management functions are adversely impacted. Ethernet Link Layer Discovery Protocol (LLDP) snooping helps identify network topology with automated network discovery, enabling better planning and monitoring. And for enhanced optical security, the T100 has line-side OTN layer 1 encryption using Advance Encryption Standard (AES-256) for each 100Gb client.
  • The new 1FINITY T400 Transport blade provides a hyper-dense architecture that is purpose-built for layer 1, 10GbE to 100G aggregation. The platform’s open optics are comprised of 10 x QSFP28 100G-BaseR service ports and 25 x QSFP+ 40 GbE client ports. A 4:1 optical fan-out cable is used to deliver up to 100 x 10GbE client ports. Like the T100, the T400 has zero touch provisioning and AC or DC power supplies. The T400 reduces energy costs with a low power consumption of 0.45W per GbE.

“Our data center interconnect solutions are designed to meet the needs of the country’s largest service providers and data center operators,” said Chris Podraza, head of the optical business unit, Fujitsu Network Communications, Inc. “1FINITY solutions offer unmatched openness, flexibility, density and security. Implementing 1FINITY products, like the latest generation of the T100 and the T400, allow customers to grow their business and to be well-positioned to meet future network needs.”

GTT Launches Global Optical Transport Services

GTT Communications, a global cloud networking provider that in January completed the acquisition of Hibernia Networks, announced the launch of Optical Transport services, designed to provide customers with scalable bandwidth and low latency connectivity for the transport of data and cloud-based applications between financial markets, data centres, media hubs and service provider networks.

The new services leverage the company's extensive subsea and terrestrial optical network, which is optimised for latency performance. The network is based on a DWDM platform and includes three owned and operated, diverse trans-Atlantic cables, as well as over 75 points of presence across North America and Europe. GTT's optical transport service features wavelengths and low latency solutions, offered at 10 and 100 Gbit/s speeds.

The company noted that the new low latency service provides high speed connectivity between major financial, media and commercial centres, delivered over GTT Express, the lowest latency trans-Atlantic cable that provides latency of less than 59 milliseconds between New York and London. The low latency service also provides card and route diversity.

Hibernia Networks and GTT announced the completion of the transaction under which GTT acquired Hibernia on January 9th. The companies originally announced the transaction in early November 2016, through which GTT was to purchase Hibernia for $590 million, including $515 million in cash and approximately 3.3 million shares of GTT common stock valued at around $75 million.

Ranovus Launches 200 Gbit/s CFP2 Direct Detect Transceiver

Ranovus, a provider of multi-terabit interconnect solutions for data centre and communications networks, announced the availability of a 200 Gbit/s CFP2 optical transceiver for data centre interconnect (DCI) applications.

The new CFP2 solution is designed to enable robust and cost-effective end-to-end scalable bandwidth connectivity for the DWDM metro network and cloud infrastructure markets where reach of 80 km and beyond is required.

The optical transceiver is based on Ranovus' technology including multi-wavelength quantum dot laser (QDL), ring resonator-based silicon photonic (SiP) modulators, driver ICs and receiver building blocks. The device also incorporates an advanced 56 Gbit/s PAM4 PHY solution with PAM4 codec and FEC capability to support a range of network applications. The 200 Gbit/s CFP2 optical transceiver is form factor-compatible with CFP2 DCO to allow the same shelf density for data centre interconnect systems.

Additional key features of the new CFP2 product include:

1. Support for DCI and metro access distances for 15, 40 and 80 km+ applications.

2. Supports 96 x DWDM channels in the C-band, plus L-band to be added in the future.

3. Provides 1.6 Tbit/s 1 RU shelf density.

4. 56 Gbit/s PAM4 PHY with multiple programmable FEC options to allow optimised link power and performance.

5. Diagnostics and self-monitoring capabilities.

Source Photonics Demos 400GBASE-FR8 OSFP module, 53 Gbaud PAM4 ROSA

Source Photonics, a provider of optical transceivers for data centre and broadband access connectivity, announced at OFC 2017 it will demonstrate for the first time its fully integrated 400GBASE-FR8 OSFP module, which it claims offers the first duplex single mode fibre module with a form factor capable of supporting a 12.8 Tbit/s 1 RU switch.

The company's 400 Gbit/s OSFP module is claimed to enable a 4x improvement in faceplate density over current 3.2 Tbit/s switches supported by 100 Gbit/s QSFP28 LR4, CWDM4 and PSM4 modules. The module integrates 8 transmit and receive channels, each operating at 50 Gbit/s using 25 Gbaud PAM4 modulation on both the electrical and the optical interfaces.

This demonstration will consist of an OSFP chassis with integrated 8-wavelength TOSA and ROSA components operating in electrical loop back through the OSFP connector with a display of the eight transmit wavelengths on an optical spectrum analyser. The prototype is based on a compact TOSA housing all eight EML lasers and wavelength multiplexing optics, as well as development of a compact ROSA package housing eight PD chips, TIAs and wavelength de-multiplexing optics.

Source Photonics also announced that at OFC, in cooperation with Semtech, a supplier of analogue and mixed-signal semiconductors, it is demonstrating a 53 GBaud PAM4 compact ROSA capable of supporting a single lambda 100 Gbit/s QSFP28.

The demonstration comprises an optical loop back from a Source Photonics' compact 53 GBaud TOSA using a production-grade EML laser and 53 GBaud ROSA based on Semtech's GN1089 TIA. The combined TOSA and ROSA link can support a loss budget of 5dB and achieve better than 1 x 10-5 BER over the tested link.

The 5G Imperative

5G was front and centre at this year's Mobile World Congress, which concluded on March 2nd. Last year, 5G was already on the tip of everyone's tongue, while this year 5G banners were all over Barcelona, from the arrival lounge at the airport to nearly every vendor stand. There is considerable industry pressure to move as quickly as possible to bring 5G to market and MWC 17 kicked off with a call by major mobile network operators and vendors to accelerate the 5G New Radio (NR) standardisation schedule to enable large-scale trials and deployments as early as 2019, a year earlier than the previous expected timeline.

The first 3GPP 5G NR specification will be part of Release 15, the global 5G standard that will make use of both sub-6 GHz and mmWave spectrum bands. An accelerated rollout for 5G could provide a strategic market advantage to certain carriers who gain a first-to-market advantage with new applications such as autonomous vehicles and fixed wireless access for residential services. Early deployments would also be welcomed by major vendors, many of which are under increasing financial pressure as new LTE installations dry up.

However, there is also a growing back pressure from some quarters, notably European operators and some regulators, not to rush things for the sake of boastfulness at the expense of technology maturity or market reality. In particular, the CTO of Telefonica was quoted in the industry press as saying that a premature lockdown of 5G specifications might prevent the technology from developing to its full potential.

Mobile operators racing to be first

In the highly competitive U.S. market, key players in 5G include Verizon, and perhaps to a slightly lesser extent AT&T. For Verizon, it is essential that it retains a network advantage over Sprint and T-Mobile. Sprint's current advertising campaign proclaiming that its network quality is within 1% of Verizon's is painful enough to the market leader, and falling behind with 5G would be disastrous. Ahead of MWC 17, Verizon announced plans to rollout 5G pre-commercial services to select customers in 11 markets by mid-2017 to create the 'largest 5G proving ground in the world'. The rollout will include several hundred cell sites that cover several thousand customer locations, with pilot markets including Ann Arbor, Atlanta, Bernardsville (New Jersey), Brockton (Massachusetts), Dallas, Denver, Houston, Miami, Sacramento, Seattle and Washington DC. Verizon's 5GTF ecosystem partners include Ericsson, Intel, Qualcomm Technologies and Samsung.

For AT&T, the first 5G business customer trial is already underway in Austin, Texas. The pilot, being conducted in partnership with Intel and Ericsson, uses millimetre wave (mmWave) technology, which can deliver multi-gigabit speeds using unlicensed spectrum. The carrier has previously reported 5G lab trials delivering speeds of up to 14 Gbit/s with less than 3 milliseconds of latency. In the first half of 2017, AT&T plans to conduct a trial in Austin, where residential customers can stream DIRECTV NOW video service over a fixed wireless 5G connection. The trial will include next-generation entertainment services over fixed 5G connections and is designed to evaluate how fixed wireless mmWave technology handles heavy video traffic.

Meanwhile in Australia, for Telstra being first to market with leading technology has also become a hallmark of the company, and Telstra was among the major carriers calling to accelerate the 5G NR standardisation schedule. Telstra has launched commercial gigabit LTE service is select capital cities across Australia enabled by LTE Advanced features including 4 x 4 MIMO, 3 CA (Carrier Aggregation) and higher order modulation (256QAM). Ericsson is a key technology provider for the rollout, with Qualcomm also a key partner.

For Korean carriers, including SK Telecom and KT, both of whom were early movers with LTE, one could say there is a national imperative to lead in 5G. The next winter Olympic Games, scheduled for February 2018, in PyeongChang, Korea, are supposed to showcase commercial 5G service. KT has previously announced plans for commercial 5G in 2019, a year earlier than others, and is also a 5G development partner with Verizon, along with vendors Ericsson and ZTE. SK Telecom likewise has announced plans for 5G NR field trials in the second half of 2017 with the goal of showcasing the technology at the Winter Games. These upcoming trials will employ 3GPP 5G NR MIMO antenna technology with adaptive beamforming and beam tracking techniques, including non-line-of-sight (NLOS) environments and device mobility. Vendor partners again include Qualcomm and Ericsson.

In Japan, NTT DOCOMO has previously stated its intention to roll out commercial 5G services across Japan ahead of the 2020 Olympic Games in Tokyo. At MWC 17, NTT DOCOMO announced interoperability testing and over-the-air field trials in Japan based on the 5G NR specifications being developed by 3GPP. The trials will operate in mid-band spectrum at 4.5 GHz, as well as mmWave spectrum at 28 GHz, showcasing the unified 5G NR design across diverse spectrum bands. The trials will utilise device prototype and base station solutions from Qualcomm Technologies and Ericsson, respectively, along with trial environments from NTT DOCOMO to simulate real-world scenarios across a broad set of use cases and deployment scenarios.

DOCOMO said It is looking forward to timely commercial network launches based on 3GPP Release 15 standard-compliant 5G NR infrastructure and devices. The trial will showcase advanced 3GPP 5G NR technologies including MIMO antenna technology, beamforming techniques, adaptive self-contained TDD, scalable OFDM-based waveforms to support wider bandwidths, advanced coding and modulation schemes and a new flexible, low-latency slot structure based design. In addition, the trial will feature 5G NR operation in mmWave spectrum at 28 GHz, employing advanced 5G NR antenna technology to deliver robust and sustained mobile broadband communications including in non-line-of-sight (NLOS) environments and device mobility.

Vendors race to be first

Among vendors, there are also bragging rights and strategic imperatives at play, For Ericsson, 5G is the really the big bet that is keeping the company alive. The Ericsson stand at MWC was mostly a showcase for its 5G ambitions, and for Ericsson this future cannot come fast enough. The company's recent financial performance and on-going restructuring are no secret, with Q4 2016 revenue down 11% compared to a year earlier and the company reporting a climate of weak investments in mobile broadband in most regions of the world with many legacy 4G projects having wrapped up. At MWC 17, Ericsson showcased a 5G platform comprising the 5G core, radio and transport portfolios, together with digital support systems, transformation services and security, all clearly aimed at mobile operators seeking to be first movers in their markets.

For Nokia, the 5G imperative story is pretty much the same, although the company benefits from a more diverse product portfolio that includes IP and optical platforms that will sustain the company in the event that a 5G rollout is delayed or more gradual than expected. At MWC 17, Nokia also said it intends to give operators a first-to-market advantage based on early specifications. The company announced its 5G FIRST end-to-end solution incorporating its AirScale and AirFrame technology, including AirScale massive MIMO Adaptive Antenna, Cloud Packet Core and mobile transport. The solution is expected to launch in the second half of 2017.

For Huawei and ZTE, 5G is the opportunity to pull ahead of their western counterparts by being first to market. At MWC, ZTE demonstrated a range of 5G mmWave and sub-6 GHz pre-commercial base stations supporting 3GPP's 5G NR new air interfaces and mainstream 5G frequency bands. The base stations use massive MIMO, beam tracking, beamforming and other key 5G technologies to achieve a 50 Gbit/s peak rate. ZTE completed its first 5G mmWave field trials last year. ZTE also showed off an upcoming 5G-capable Gigabit Phone powered by the Qualcomm Snapdragon 835 chipset and which combines wireless carrier aggregation with 4 x 4 MIMO antenna technology and 256QAM modulation. Some may say this device is really pre-5G because it uses technology that is becoming available for 4G LTE, yet the announcement indicates that ZTE believes 5G handsets will be crucial for rapid upgrades to 5G infrastructure.

The issue of 5G handsets and other devices leads to Qualcomm, for whom it would simply be unacceptable to lose its lead. Anyone walking in to Barcelona's Fira convention centre saw dozens of banners proclaiming 'Snapdragon Gigabit LTE - first place in the race'. A rapid adoption of 5G NR would certainly be welcome by the California-based silicon vendor. In October 2016 Qualcomm made the first commercial 5G modem chipset announcement with its Snapdragon X50 5G modem, clearly aimed at operators and OEMs conducting lab tests, field trials and early deployments. The Snapdragon X50 5G modem initially supports operation in mmWave spectrum in the 28 GHz band and employs MIMO antenna technology with adaptive beamforming and beam tracking techniques for NLOS environments. It also offers 800 MHz bandwidth support for peak download speeds of up to 5 Gbit/s.

At MWC 17, Qualcomm announced support for the 5G NR accelerated plan and is expanding its Snapdragon X50 5G modem family to include 5G NR multi-mode chipset solutions compliant with the 3GPP-based 5G NR global system supporting operation in the sub-6 GHz and multi-band mmWave spectrum. This includes support of both Non-Standalone (NSA) operation (where control signalling is sent over LTE), and Standalone (SA) operation (where all control signalling and user data are sent over 5G NR), and are designed to enable the next generation of premium-tier mobile cellular devices while also aiding operators to execute early 5G trials and deployments. The first commercial products integrating 5G NR modems from the Snapdragon X50 family are expected to be available to support the first large-scale 5G NR trials and commercial network launches starting in 2019.

Big questions remain

The question remains, is there a business case for an accelerated adoption of 5G? The upgrade to 4G brought with it concurrent improvements to the overall network infrastructure; 5G is being preceded with network advancements, including network virtualisation technologies, open systems, and even continued improvements of the underlying metro optical transport network. In addition, 4.5G technologies such as carrier aggregation and MIMO are already adding significant capacity to mobile networks. The chief financial officers of the mobile network operators will look at this and ask whether there are in fact end customers willing to pay to be the first aboard fully 5G commercial services.

Perhaps there are certain customers, like the operators of fully-autonomous vehicle fleets, that are eager to sign up for the first available 5G service, but for now it looks as though certain vendors and operators are pressing full-steam ahead because they have no other choice than to be first with 5G.

OE Solutions Intros Smart SFPs for TDM over Packet

South Korea's OE Solutions and AimValley of the Netherlands have announced the introduction of the final products within their portfolio of Network Migration Smart SFPs, which enable customers to deliver TDM services over packet network via a plug-and-play portfolio that offers support for TDM services from 1.5 Mbit/s up to 2.5 Gbit/s.

The family of network migration Smart SFP modules is designed to facilitate the transition from legacy TDM to all-packet networks by converting PDH or SONET/SDH signals to a packet stream.

The Smart SFP products are designed to provide operators with a cost-effective alternative to upgrading systems using dedicated TDM-over-packet cards or installing new equipment. The Smart SFP modules can be added to any router or packet switch where a TDM service is required.

OE Solutions stated that adding legacy PDH or SDH/SONET services to a packet network can be achieved via first provisioning an EPL service into the packet network and then inserting a Smart SFP at each endpoint of the EPL service. This provides a reliable point to point connection for the delivery of services such as E1 to a legacy base station.

For more complex application, such as PDH to SONET aggregation, Smart SFPs are inserted at locations where an PDH or SONET service is required, and using EPL or EVPL the connections in the network can be set up to aggregate legacy services, for example T1 service aggregation into an OC-3 service. The companies noted that Smart SFPs support industrial temperature range operation and offer a range of optical and electrical interfaces for different applications.

Last week, OE Solutions announced the release of extended reach 10 Gbit/s CWDM SFP+ transceivers and high-density transceivers, including 2 x 10 Gbit/s compact SFP+, 4 x 10 Gbit/s QSFP+ and 25 Gbit/s SFP28, for wireless fronthaul applications.

The CWDM SFP+ transceivers support 1 and 10 Gigabit Ethernet, as well as CPRI-2 to CPRI-8 for wireless network applications, and utilise cooled DFB laser diodes combined with wide dynamic range APD detectors to enable reach up to 20 km for wavelengths up to 1611 nm.

The new industrial temperature grade, high density transceivers are designed to enable high port densities in radio access network equipment to support the roll-out of next generation wireless infrastructures. Products include a 2 x 10 Gbit/s multi-rate bidirectional compact-SFP+ solution for distances up to 20 km, a 4 x 10 Gbit/s multi-rate QSFP+ device offering reach up to 300 metres, and a 25 Gbit/s SFP28 solution with multiple rate and protocol options for distances up to 10 km.