It is interesting to think about the amount of data that is currently being transmitted which is now enabled by Cloud networks. Every day people watch hundreds of millions of hours on YouTube and generate billions of views.1 One third of all data will soon be stored in the Cloud, as reported by Gartner.2 And, according to the Cisco Global Cloud Index, by 2018, 59% of the total cloud workloads will be Software-as-a-Service (SaaS) workloads, up from 41% in 2013.3
In a world where Cloud services are growing quickly and high capacity connectivity is paramount, networks need to be scalable, simplified and flexible. This topic was top of mind in the presentations and discussions at Transport Networks Russia 2015, a conference focused on the development of telecommunications transport networks in Russia and CIS.
In the wake of an ever-visible shift toward SDN-enabled infrastructures, maximum efficiency and convergence of the network layers seems eminent. Aligned with this thought process, Infinera premiered two new photonic integrated circuits to support the emerging Layer C (Cloud Services) and Layer T (Transport) network model. Network Function Virtualization (NFV) provides a means to migrate network functions related to the upper layers of the network from dedicated appliances to software services on commercial x86 servers within Cloud data centers; leading to the creation of the Cloud services layer or Layer C. In order to support Layer C, Cloud datacenters and end users need to be interconnected by a highly scalable, converged and flexible transport network, or Layer T.
At TransNet Russia, the participants discussed various approaches to build Layer T. All agreed that there’s a continuing need to provide scalability. I believe that the foundation of scale is integrated photonics, which provides more capacity per line card and system while simplifying the network – fewer boxes, fibers and modules, and less space, power and fewer manual processes. Photonic Integrated Circuits (PICs) are integral to the evolution of the transport network providing significant benefits when integrated into a packet-optical DWDM transport system for an efficient Layer T, ultimately allowing Layer C to thrive.
Service providers in the region have to support the prevailing tough economic conditions and are required to perform against rigorous financial metrics set by shareholders. At the same time they have to stay on top of the competition while consumer demands for capacity and bandwidth skyrocket. As part of Intelligent Transport Network philosophy, Infinera’s goal is to help providers drive expenses down and at the same time keep pace with network growth. Recently Infinera announced two new PICs, ePIC-500 and oPIC-100, to bring the economics of the integrated photonics for the metro.
Regardless of topology (hub-and-spoke, mesh or ring), with the new PICs networks showed an estimated average reduction of 28 percent in modules, 31 percent in power and 45 percent in bandwidth inefficiencies as compared to conventional technologies that deliver 100G, 200G or 400G. Finer granularity enabled by new PICs and tight integration of optical components on the chip results in unmatched flexibility of architectures and significant savings in CapEx and OpEx – something that the Russian operators are keenly focused upon and as exemplified in the key session of “Expansion of Transport Networks” at TransNet 2015.
As Cloud services continue to develop, it is very important that transport networks are scalable yet granular and simple to operate. Sliceable photonics provided by the Infinera ePIC-500 and oPIC-100 are designed to allow operators to lower the costs and complexities of transport while handling enormous traffic growth. As one participant at TransNet 2015 Russia quipped “It was one of the most famous and important conferences for Russian communications companies on the new and complicated stage of market development, and it became intense, relevant and practical”. Infinera is glad to be a part of and lead the Russia market development.
To learn more about Infinera’s sliceable photonics click here.
- Cisco Global Cloud Index: Forecast and Methodology, 2013–2018
By Geoff Bennett, Director, Solutions and Technology
The show I will always think of as “WDM”, but which now rejoices in the snappy title of “WDM and Next Generation Optical Networking” took place recently in Nice, France.
All the things you’d expect were there – sun drenched Mediterranean beaches, superb French cuisine…and a major taxi driver strike in protest against Uber! It wouldn’t be France without at least one national strike that causes chaos for the conference attendees. Check out this graphic!
Another thing that was no surprise was the focus of the first day of the event – the increasing demands of a Cloud Services Layer (Layer C) driving the need for scalability, flexibility and programmability in the underlying transport network (Layer T). These dynamics are well described in the Day One Sessions in which I took part.
We heard from Andrew Schmitt of IHS on the growth of capacity in agreement with his predictions of the Long Haul Optical Reboot, and that this trend towards 100G is already taking off in the metro.
Amongst the other presentations in the morning we heard how BT and TeliaSonera International Carrier have been taking advantage of coherent technologies to increase fiber capacity.
In particular, Kevin Smith of BT outlined three axes of capacity scaling – the first axis is increasing the Baud rate of individual wavelengths, although this is an axis that is severely limited by the performance of today’s opto-electronics. The second axis is increasing the number of bits carried by each symbol, in other words the move to higher order modulation such as 8QAM and 16QAM. We have several blogs (The 8QAM Sweet Spot, Fire Up The Flux Capacity and Crank It Up To Terabit Scale, Infinera’s Expanding FlexCoherent Toolkit allows Subsea Operators to Maximize Fiber Capacity and Reach) on this topic on the Infinera site already, and there seems to be general acceptance that 16QAM will not be a universal solution for optical transmission over today’s fiber.
(As a side note, Ian Redpath of Ovum presented the results of his informal survey of conference attendees in which less than 50% of the respondents seem to be “sold” on 16QAM at the moment).
Kevin’s third axis is the move to super-channels and in his case, he presented the evolution of the conventional DWDM vendors who have to buy expensive optical components for single-wavelength implementations (even if multiple circuits are implemented on one line card). These approaches tend to exhaust the increase in Baud rate and modulation order first.
This is where Infinera’s mastery of large scale PIC technology makes a pivotal difference. Large scale (ie. multiple wavelengths on one chip) PICs make super-channels practical today – in a way that discrete optics (and pluggables) simply cannot. Since the first commercial super-channel shipments in 2012, PICs have given Infinera the ability to radically differentiate our solution.
This was the focus of my presentation, where I showed the commercial success of super-channels in the long haul and then posed the question – could they achieve the same success in metro? Ironically, the word “metro” in this context covers a lot of ground (i.e. the metro market is made up of multiple segments), and we have certainly seen the success of high capacity, super-channel solutions in one of these segments – the Data Center Interconnect market. Several speakers acknowledged that this is a market that has essentially been created by Infinera last year with our introduction of the Cloud Xpress, and is now being copied with several announced (but not yet shipping) products from other DWDM vendors.
Another, distinct segment of “metro” is metro aggregation. Here we tend to see high capacity hub sites that certainly will benefit from super-channel operational scalability (i.e. install the line card and forget about it), but these sites aggregate traffic from spoke sites that are probably prime candidates for a move to 100Gb/s connection speeds.
This is the application area targeted by Infinera’s “sliceable PIC” technology. Above you can see how this works. Our sliceable PIC is located in the hub site, and FlexROADM technology is used to slice its capacity out to multiple spoke sites. At these locations, we deploy the application-optimized 100G PIC (oPIC-100). Service providers can also use Instant Bandwidth™ to turn on new super-channel capacity without visiting that location. The only truck roll needed is to install the oPIC-100 units in the spoke sites.
The combination of these technologies delivers operational scalability, cash flow optimization, and bandwidth efficiency.
And “metro” is still a much wider market than these two segments. We have metro access and metro packet as distinct segments, as well as mobile backhaul and mobile fronthaul.
Those of you who have been following the news will be aware that Infinera has made a commercial offer to acquire the Swedish company, Transmode. Their product range aligns perfectly with the existing Infinera solution set, with just enough overlap to allow a seamless, end to end solution to address the whole of the metro, long haul and subsea markets.
It’s clear that, in order to thrive, Layer C has to be “fed” by a scalable, dynamic and efficient Layer T. PIC-based super-channels have already proven themselves to be the perfect Layer T technology in the long haul market. It turns out that the same kind of point to point super-channel is an ideal solution for a simple Data Center Interconnect applications too. But now the addition of sliceability allows us to deploy PIC technology that’s ideal for the metro aggregation segment.
So, all in all, I have to say this was the most interesting WDM-NGON conference I’ve attended for many years – and not just because of the taxi strike!
Infinera’s Cloud Xpress continues to gain momentum – first as a Leading Lights Award finalist in the “Best New Product (Telecom)” category, then with the expansion of our CX Product Family (up to 100 GbE for high capacity data center interconnect) and now with the “Best of Show Award” at Interop 2015 in Tokyo, Japan.
Interop, one of the world’s leading independent technology conferences, is designed to inspire, inform and connect the world’s IT community. This year, Interop Japan had more than 132,000 attendees learning about the latest advancements in networking, virtualization, cloud, mobility and data centers.
We were thrilled to have Hideki Ameku, Sr. Systems Engineer, APAC Sales at Infinera, along with our channel partner Nissho Electronics Japan, provide live demonstrations of the Cloud Xpress, showing its simplicity, scale and ease of use, during the event.
Industry experts engaged in several rounds of stringent assessments and the judges were impressed, calling it “the next generation of WDM.” Nissho Electronics was awarded the 2015 Best of Show Award in the data center networking category for Infinera’s Cloud Xpress.
It was selected for its power efficiency at 1 W/Gb/s and Instant Bandwidth™ capability, enabling rapid deployment of bandwidth in 100 Gb/s (100G) increments without having to deploy any additional hardware.
The Cloud Xpress family of Metro Cloud platforms offers a complete solution – Hyper-Scale Density + Simplified Operations + Low Power– and is designed specifically to address the needs of Cloud service providers, Internet content providers, Internet Exchange service providers, large enterprises and other large-scale datacenter operators. We carved out this market in 2014 using input directly from customers, and are proud to be the only company shipping a compact rack-and-stack form factor: 1Tb/s of Input/Output in a short 2 RU (3.75”). We’ve seen strong interest from customers and have been shipping Cloud Xpress platforms since December 2014.
And, just last month, we announced the expansion of the Cloud Xpress family with a new Cloud Xpress platform that is designed to interconnect higher capacity 100 GbE client devices while expanding our feature set to include – including MACsec encryption, support for NETCONF & YANG, SDN protocols that enable automation and LLDP discovery protocols, that enable discovery of Ethernet Switches & Routers across the wide area network.
The “Best of Show” award is not only a testament to our innovation, but also to our channel partnerships that help us drive mutual value (and mindshare) with our customers. We look forward to meeting (and exceeding) customer demands both directly and through our channels as we continue to advance the metro Cloud market.
Learn more about Infinera’s Cloud Xpress Product Family.
Fiber Network Optimization: Get More from your Fiber with Multi-Carrier Optical Super-Channels – A Perspective from ANGA COM
By Gaylord Hart, Director MSO Market Segment
I recently attended the ANGA COM Exhibition and Congress in Cologne, Germany. With 17,000+ participants from 74 countries, 450 exhibitors and 150 speakers, this conference continues to be a popular place for the international cable telecommunications community to learn and share business and technical challenges and solutions.
While there, I, along with other distinguished panelists, discussed various ways to improve the fiber plant, including access & core networks. The focus of my presentation was on terabit super-channels and how these can be implemented flexibly and cost-effectively while increasing fiber spectral capacity.
Implementing a terabit super-channel with just one or two waves is challenging due to constraints on commercially available high-speed electronics that would be required to process 192 or 384 gigabaud speeds. Instead, the fastest time to market will likely use a multi-carrier approach. If the constituent waves in the super-channel are based on 100G technology, implementing such super-channels is readily possible with existing technology.
However, there is a tradeoff here. A 1.2 Tb/s super-channel, for example, would require 12 x 100G waves in the super-channel and represents a significant increase in component count and complexity, especially if implemented using discrete photonics. Fortunately, modern photonic integration technology solves this problem. Since a photonic integrated circuit (PIC) combines multiple discrete optical components as well as their optical interconnects on a single monolithic chip that is manufactured in a single cycle, most of the complexity and cost issues associated with multi-carrier super-channels goes away. What you also end up with is better density and reliability, but less heat and lower power consumption.
But perhaps the greatest advantage to multi-carrier super-channels is the flexibility they offer. By allowing each wave in the super-channel to be individually tuned and provisioned for modulation format, the 12 channel 1.2 Tb super-channel used in the above example could be configured as a single super-channel (1 x 1.2 Tb), three super-channels (3 x 400G), 12 individual channels or any combination of these. This allows the individual super-channels or waves to be independently routed through the network, where bandwidth is needed most and provisioned for modulation format to optimize reach versus bandwidth as needed.
To learn more, check out these resources:
- White Paper: The Evolution of Next-Gen Optical Networks – Terabit Super-Channels and Flexible Grid ROADM Architecture
- Video: Industry’s First Single Card Terabit Field Trial
- Video: Infinera DTN-X Chassis Upgrade
- Terabit Scale Super-Channels
By Jay Gill, Principal Product Marketing Manager at Infinera
With the launch of the Cloud Xpress family in 2014, Infinera created a new market for cloud-optimized, hyper-scale datacenter transport. From the beginning, we said that Cloud Xpress would support customers needing to interconnect datacenters with 10 Gigabit Ethernet (GbE), 40 GbE and 100 GbE. Last week, we made good on the promise, announcing the newest Cloud Xpress with 100 GbE to complement the platforms optimized for 10 and 40 GbE, which have been shipping since late 2014.
Some observers asked us, “Does the market really need 100 GbE yet, or is this going to be a niche product for a while?”
Our answer is clear: The time for 100 GbE is now and we plan to sell a lot of it.
I understand where the question is coming from. Though 100 GbE has been hyped for some time, early implementations were expensive, bulky and power-hungry. But as with every prior transition between Ethernet data rates, 100 GbE has clearly reached the tipping point. 100 GbE ports on datacenter switches are now more cost-effective and provide better density than 10 or 40 GbE for anyone who needs the bandwidth. QSFP28 optical modules support 100 GbE in the same form factor as the popular 40 GbE QSFP+, and at a much higher density than alternatives such as CFP or CFP2 modules.
As a result, 100 GbE ports using QSFP28 are taking off inside the datacenter. Based on various analyst projections, it looks likely that 100G datacenter switch port shipments will surpass 40G within 2 years and comprise over a third of all new ports by 2018. But those aggregate statistics hide the fact that hyper-scale datacenter operators, such as Internet Content Providers (ICPs) and Internet Exchange (IX) providers, are leading the 100G adoption and moving much faster. They are already telling us that they have an urgent need to extend 100 GbE links between their datacenters this year.
That’s why Cloud Xpress with 100 GbE is perfectly timed for the market.
And, as far as we can tell, Cloud Xpress will also be the first optical transport product of its class shipping 100 GbE ports based on QSFP28.
The density of QSFP28 allowed us to include five 100 GbE ports and still have room for sixteen 10 GbE SFP+ ports on the narrow 2 RU faceplate. This allows customers to fill up the 500 Gb/s line capacity with a flexible mix of 10 and 100 GbE links – without any breakout cables — which can be useful in many growth and migration scenarios.
As we look ahead, we can see even greater 100 GbE densities aligning with increasing optical line capacity provided by Infinera’s unique PIC technology, which will allow the Cloud Xpress family to continue to scale along with our customers’ ever-growing bandwidth demands. Come along and ride the 100G wave!
For more information, check out the following links:
- Press Release: Infinera Expands Cloud Xpress Family for High Capacity Datacenter Interconnect
- Video: Unleashing Cloud Networks
- Case Study: ICPs Now Choose Metro Cloud Optimized Appliance for Networks Connectivity
- Website: Infinera’s Cloud Xpress Family
By Mark Showalter
While much of the optical industry is enjoying the 17th Annual Next Generation Optical Networking show in Nice this week, I am basking in the light of the recently published 2015 Optical Network Hardware Vendor Scorecard from Andrew Schmitt, research director, Carrier Transport Networking at IHS (formerly Infonetics Research). In the report IHS names Infinera the ‘ne plus ultra of the rapidly growing Internet content provider and datacenter interconnection market’.
Latin is not my strong suit so I referred to the Cloud to look up what this meant and found the following from Merriam-Webster:
Definition of NE PLUS ULTRA
1 : the highest point capable of being attained
2 : the most profound degree of a quality or state
Hard to top that! The report further highlights Infinera as leading with ‘outstanding customer perceptions, large market share gains and tight finances.’
Last year, Infonetics Research ranked Infinera the top optical vendor in the 2014 Optical Network Hardware Vendor Scorecard, reporting “Infinera is #1 this year as a result of very high ratings among customers for technology innovation, reliability, and service and support.”
This year, IHS (formerly Infonetics Research) refined their scoring system and ranked each of the top ten optical hardware vendors into one of three categories: Leader, Established or Challenger, depending on overall score. We are thrilled to see Infinera named as one of three to make up the top category – Leader.
Andrew Schmitt offered further insight into the ranking: “Infinera has excellent scores across the board and as such it is no surprise that Infinera is ranked as a leader. Infinera has market presence with excellent scores from customers for product reliability, service and support. The company also had excellent momentum as a result of big market share gains in 2014 and is the top perceived vendor when it comes to technology innovation.”
While all of us at Infinera are delighted by this award, I think our CEO Tom Fallon says it best:
“The IHS Optical Network Hardware Vendor scorecard shows that Infinera’s focus on customer satisfaction, quality and market leading innovation continues to resonate with global network operators. We are honored to receive the highest recognition from IHS for the second year running based on our success deploying Intelligent Transport Networks around the world.”