An insider's view on the merits of optical integration
Tolstikhin is president and CEO of Intengent, the Ottawa-based consultancy and custom design service provider, and an industry veteran of photonic integration. In 2005 he founded OneChip Photonics, a fabless maker of indium phosphide photonic integrated circuits for optical access.
One important lesson he learned at OneChip was how the cost benefit of a photonic integrated circuit (PIC) can be eroded with a cheap optical sub-assembly made from discrete off-the-shelf components. When OneChip started, the selling price for GPON optics was around $100 a unit but this quickly came down to $6. "We needed sales in volumes and they never came close to meeting $6," says Tolstikhin.
OneChip changed strategy, seeing early the emerging opportunity for 100-gigabit optics for the data centre but despite being among the first to demonstrate fully integrated 100-gigabit transmitter and receiver chips – at OFC 2013 – the company eventually folded.
When OneChip started, the selling price for GPON optics was around $100 a unit but this quickly came down to $6
Integent can be seen as the photonic equivalent of an electronic ASIC design house that was common in the chip industry, acting as the intermediary between an equipment vendor commissioning a chip design and the foundry making the chip.
Integent creates designs for system integrators which it takes to a commercial foundry for manufacturing. The company makes stand-alone devices, device arrays, and multi-function PICs. Integent uses the regrowth-free taper-assistant vertical integration (TAVI) indium phosphide process of the California-based foundry Global Communication Semiconductors (GCS). "We have also partnered with a prominent PIC design house, VLC Photonics, for PIC layout and verification testing,” says Tolstikhin. Together, Intengent, VLC and GCS offer a one-stop-shop for the development and production of PICs.
III-V and silicon photonics
Tolstikhin is a big fan of indium phosphide and related III-V semiconductor materials, pointing out that they can implement all the optical functions required for telecom and datacom applications. He is a firm believer that III-V will continue to be the material system of choice for various applications and argues that silicon photonics is not so much a competitor to III-V but a complement.
"Silicon photonics needs indium-phosphide-based sources but also benefits from III-V modulators and detectors, which have better performance than their silicon photonics counterparts," he says.
He admits that indium phosphide photonics cannot compete with the PIC scalability that silicon photonics offers. But that will benefit indium phosphide as silicon photonics matures. Intengent already benefits from this co-existence, offering specialised indium phosphide photonic chip development for silicon photonics as well.
"Silicon photonics cannot compete with indium phosphide photonics in relatively simple yet highest volume optical components for telecom and datacom transceivers," says Tolstikhin. Partly this is due to silicon photonics' performance inferiority but mainly for economical reasons.
Silicon photonics will have its chance, but only where it beats competing technologies on fundamentals, not just cost
There are also few applications that need monolithic photonic integration. Tolstikhin highlights coherent optics as one example but that is a market with limited volumes. Meanwhile, the most promising emerging market - transceivers for the data centre, whether 100-gigabit (4x25G NRZ) PSM or CWDM4 designs or in future 400-gigabit (4x100G PAM4) transceivers, will likely be implemented using optical sub-assembly and hybrid integration technologies.
Tolstikhin may be a proponent of indium phosphide but he does not dismiss silicon photonics' prospects: "It will have its chance, but only where it beats competing technologies on fundamentals, not just cost."
One such area is large-scale optoelectronic systems, such as data processors or switch fabrics for large-scale data centres. These are designs that cannot be assembled using discretes and go beyond the scalability of indium phosphide PICs. "This is not silicon photonics-based optical components instead of indium phosphide ones but a totally different system and possibly network solutions," he says. This is also where co-integration of CMOS electronics with silicon photonics makes a difference and can be justified economically.
He highlights Rockley Photonics and Ayar Labs as start-ups doing just this: using silicon photonics for large-scale electro-photonic integration targeting system and network applications. "There may also be more such companies in the making," says Tolstikhin. "And should they succeed, the entire setup of optics for the data centre and the role of silicon photonics could change quite dramatically."
SDM and MIMO: An interview with Bell Labs
Part 2: The capacity crunch and the role of SDM
The argument for spatial-division multiplexing (SDM) - the sending of optical signals down parallel fibre paths, whether multiple modes, cores or fibres - is the coming ‘capacity crunch’. The information-carrying capacity limit of fibre, for so long described as limitless, is being approached due to the continual yearly high growth in IP traffic. But if there is a looming capacity crunch, why are we not hearing about it from the world’s leading telcos?
“It depends on who you talk to,” says Peter Winzer, head of the optical transmission systems and networks research department at Bell Labs. The incumbent telcos have relatively low traffic growth - 20 to 30 percent annually. “I believe fully that it is not a problem for them - they have plenty of fibre and very low growth rates,” he says.
Twenty to 30 percent growth rates can only be described as ‘very low’ when you consider that cable operators are experiencing 60 percent year-on-year traffic growth while it is 80 to 100 percent for the web-scale players. “The whole industry is going through a tremendous shift right now,” says Winzer.
In a recent paper, Winzer and colleague Roland Ryf extrapolate wavelength-division multiplexing (WDM) trends, starting with 100-gigabit interfaces that were adopted in 2010. Assuming an annual traffic growth rate of 40 to 60 percent, 400-gigabit interfaces become required in 2013 to 2014, and the authors point out that 400-gigabit transponder deployments started in 2013. Terabit transponders are forecast in 2016 to 2017 while 10 terabit commercial interfaces are expected from 2020 to 2024.
In turn, while WDM system capacities have scaled a hundredfold since the late 1990s, this will not continue. That is because systems are approaching the Non-linear Shannon Limit which estimates the upper limit capacity of fibre at 75 terabit-per-second.
Starting with 10-terabit-capacity systems in 2010 and a 30 to 40 percent core network traffic annual growth rate, the authors forecast that 40 terabit systems will be required shortly. By 2021, 200 terabit systems will be needed - already exceeding one fibre’s capacity - while petabit-capacity systems will be required by 2028.
Even if I’m off by an order or magnitude, and it is 1000, 100-gigabit lines leaving the data centre; there is no way you can do that with a single WDM system
Parallel spatial paths are the only physical multiplexing dimension remaining to expand capacity, argue the authors, explaining Bell Labs’ interest in spatial-division multiplexing for optical networks.
If the telcos do not require SDM-based systems anytime soon, that is not the case for the web-scale data centre operators. They could deploy SDM as soon as 2018 to 2020, says Winzer.
The web-scale players are talking about 400,000-server data centres in the coming three to five years. “Each server will have a 25-gigabit network interface card and if you assume 10 percent of the traffic leaves the data centre, that is 10,000, 100-gigabit lines,” says Winzer. “Even if I’m off by an order or magnitude, and it is 1000, 100-gigabit lines leaving the data centre; there is no way you can do that with a single WDM system.”
SDM and MIMO
SDM can be implemented in several ways. The simplest way to create parallel transmission paths is to bundle several single-mode fibres in a cable. But speciality fibre can also be used, either multi-core or multi-mode.
For the demo, Bell Labs used such a fibre, a coupled 3-core one, but Sebastian Randel, a member of technical staff, said its SDM receiver could also be used with a fibre supporting a few spatial modes. By increasing slightly the diameter of a single-mode fibre, not only is a single mode supported but two second-order modes. “Our signal processing would cope with that fibre as well,” says Winzer.
The signal processing referred to, that restores the multiple transmissions at the receiver, implements multiple input, multiple output or MIMO. MIMO is a well-known signal processing technique used for wireless and digital subscriber line (DSL).
They are garbled up, that is what the rotation is; undoing the rotation is called MIMO
Multi-mode fibre can support as many as 100 spatial modes. “But then you have a really big challenge to excite all 100 spatial modes individually and detect them individually,” says Randel. In turn, the digital signal processing computation required for the 100 modes is tremendous. “We can’t imagine we can get there anytime soon,” says Randel.
Instead, Bell Labs used 60 km of the 3-core coupled fibre for its real-time SDM demo. The transmission distance could have been much longer except the fibre sample was 60 km long. Bell Labs chose the coupled-core fibre for the real-time MIMO demonstration as it is the most demanding case, says Winzer.
The demonstration can be viewed as an extension of coherent detection used for long-distance 100 gigabit optical transmission. In a polarisation-multiplexed, quadrature phase-shift keying (PM-QPSK) system, coupling occurs between the two light polarisations. This is a 2x2 MIMO system, says Winzer, comprising two inputs and two outputs.
For PM-QPSK, one signal is sent on the x-polarisation and the other on the y-polarisation. The signals travel at different speeds while hugely coupling along the fibre, says Winzer: “The coherent receiver with the 2x2 MIMO processing is able to undo that coupling and undo the different speeds because you selectively excite them with unique signals.” This allows both polarisations to be recovered.
With the 3-core coupled fibre, strong coupling arises between the three signals and their individual two polarisations, resulting in a 6x6 MIMO system (six inputs and six outputs). The transmission rotates the six signals arbitrarily while the receiver, using 6x6 MIMO, rotates them back. “They are garbled up, that is what the rotation is; undoing the rotation is called MIMO.”
Demo details
For the demo, Bell Labs generated 12, 2.5-gigabit signals. These signals are modulated onto an optical carrier at 1550nm using three nested lithium niobate modulators. A ‘photonic lantern’ - an SDM multiplexer - couples the three signals orthogonally into the fibre’s three cores.
The photonic lantern comprises three single-mode fibre inputs fed by the three single-mode PM-QPSK transmitters while its output places the fibres closer and closer until the signals overlap. “The lantern combines the fibres to create three tiny spots that couple into a single fibre, either single mode or multi-mode,” says Winzer.
At the receiver, another photonic lantern demultiplexes the three signals which are detected using three integrated coherent receivers.
Don’t do MIMO for MIMO’s sake, do MIMO when it helps to bring the overall integrated system cost down
To implement the MIMO, Bell Labs built a 28-layer printed circuit board which connects the three integrated coherent receiver outputs to 12, 5-gigabit-per-second 10-bit analogue-to-digital converters. The result is an 600 gigabit-per-second aggregate output digital data stream. This huge data stream is fed to a Xilinx Virtex-7 XC7V2000T FPGA using 480 parallel lanes, each at 1.25 gigabit-per-second. It is the FPGA that implements the 6x6 MIMO algorithm in real time.
“Computational complexity is certainly one big limitation and that is why we have chosen a relatively low symbol rate - 2.5 Gbaud, ten times less than commercial systems,” says Randel. “But this helps us fit the [MIMO] equaliser into a single FPGA.”
Future work
With the growth in IP traffic, optical engineers are going to have to use space and wavelengths. “But how are you going to slice the pie?” says Winzer.
With the example of 10,000, 100-gigabit wavelengths, will 100 WDM channels be sent over 100 spatial paths or 10 WDM channels over 1,000 spatial paths? “That is a techno-economic design optimisation,” says Winzer. “In those systems, to get the cost-per-bit down, you need integration.”
That is what the Bell Lab’s engineers are working on: optical integration to reduce the overall spatial-division multiplexing system cost. “Integration will happen first across the transponders and amplifiers; fibre will come last,” says Winzer.
Winzer stresses that MIMO-SDM is not primarily about fibre, a point frequently misunderstood. The point is to enable systems with crosstalk, he says.
“So if some modulator manufacturer can build arrays with crosstalk and sell the modulator at half the price they were able to before, then we have done our job,” says Winzer. “Don’t do MIMO for MIMO’s sake, do MIMO when it helps to bring the overall integrated system cost down.”
Further Information:
Space-division Multiplexing: The Future of Fibre-Optics Communications, click here
For Part 1, click here
Hybrid integration specialist Kaiam acquires Gemfire
Kaiam Corp. has secured US $16M in C-round funding and completed the acquisition of Gemfire.
"We have a micro-machine technology that allows us to use standard pick-and-place electronic assembly tools, and with our micro-machine, we achieve sub-micron alignment tolerances suitable for single-mode applications"
Byron Trop, Kaiam
With the acquisition, Kaiam gains planar lightwave circuit (PLC) technology and Gemfire's 8-inch wafer fab in Scotland. This is important for the start-up given there are few remaining independent suppliers of PLC technology.
Working with Oplink Communications, Kaiam has also demonstrated recently a 100 Gigabit 10x10 MSA 40km CFP module.
Hybrid integration technology
Kaiam has developed hybrid integration technology that achieves sub-micron alignment yet only requires standard electronic assembly tools.
"With single-mode optics, it is very, very difficult to couple light between components," says Byron Trop, vice president of marketing and sales at Kaiam. "Most of the cost in our industry is associated with aligning components, testing them and making sure everything works."
The company has developed a micro-machine-operated lens that is used to couple optical components. The position of the lens is adjustable such that standard 'pick-and-place' manufacturing equipment with a placement accuracy of 20 microns can be used. "If you set everything [optical components] up in a transceiver with a 20-micron accuracy, nothing would work," says Trop.
Components are added to a silicon breadboard and the micro-machine enables the lens to be moved in three dimensions to achieve sub-micron alignment. "We have the ability to use coarse tools to manipulate the machine, and at the far end of that machine we have a lens that is positioned to sub-micron levels," says Trop. Photo-diodes on a PLC provide the feedback during the active alignment.
Another advantage of the technique is that any movement when soldering the micro-machine in position has little impact on the lens alignment. "Any movement that happens following soldering is dampened over the distance to the lens," says Trop. "Therefore, movement during the soldering process has negligible impact on the lens position."
Kaiam buys its lasers and photo-detector components, while a fab make its micro-machine. Hybrid integration is used to combine the components for its transmitter optical sub-assembly (TOSA) and receiver optical sub-assembly (ROSA) designs, and these are made by contract manufacturers. Kaiam has a strategic partnership with contract manufacturer, Sanmina-SCI.
The company believes that by simplifying alignment, module and systems companies have greater freedom in the channel count designs they can adopt. "Hybrid integration, this micro-alignment of optical components, is no longer a big deal," says Trop. "You can start thinking differently."
"We will also do more custom optical modules where somebody is trying to solve a particular problem; maybe they want 16 or 20 lanes of traffic"
For 100 Gigabit modules, companies have adopted 10x10 Gigabit-per-second (Gbps) and 4x28Gbps designs. The QSFP28 module, for example, has enabled vendors to revert back to four channels because of the difficulties in assembly.
"Our message is not more lanes is better," says Trop. "Rather, what is the application and don't consider yourself limited because the alignment of sub-components is a challenge."
With the Gemfire acquisition, Kaiam has its own PLC technology for multiplexing and de-multiplexing multiple 10Gbps and, in future, 25Gbps lanes. "Our belief is that PLC is the best way to go and allows you to expand into larger lane counts," says Trop.
Gemfire also owned intellectual property in the areas of polymer waveguides and semiconductor optical amplifiers.
Products and roadmap
Kaiam sells 40Gbps QSFP TOSAs and ROSAs for 2km, 10km and 40km reaches. The company is now selling its 40km 10x10 MSA TOSA and ROSA demonstrated at the recent OFC/NFOEC show. Trop says that the 40km 10x10 CFP MSA module is of great interest to Internet exchange operators that want low cost, point-to-point links.
"Low cost, highly efficient optical interconnect is going to be important and it is not all at 40km reaches," says Trop. "Much of it is much shorter distances and we believe we have a technology that will enable that."
The company is looking to apply its technology to next-generation optical modules such as the CFP2, CFP4 and QSFP28. "We will also do more custom optical modules where somebody is trying to solve a particular problem; maybe they want 16 or 20 lanes of traffic," says Trop.
Alcatel-Lucent demos dual-carrier Terabit transmission
"Without [photonic] integration you are doubling up your expensive opto-electronic components which doesn't scale"
Peter Winzer, Alcatel-Lucent's Bell Labs
Part 1: Terabit optical transmission
Alcatel-Lucent's research arm, Bell Labs, has used high-speed electronics to enable one Terabit long-haul optical transmission using two carriers only.
Several system vendors have demonstrated one Terabit transmission including Alcatel-Lucent but the company is claiming an industry first in using two multiplexed carriers only. In 2009, Alcatel-Lucent's first Terabit optical transmission used 24 sub-carriers.
"There is a tradeoff between the speed of electronics and the number of optical modulators and detectors you need," says Peter Winzer, director of optical transmission systems and networks research at Bell Labs. "In general it will be much cheaper doing it with fewer carriers at higher electronics speeds than doing it at a lower speed with many more carriers."
What has been done
In the lab-based demonstration, Bell Labs sent five, 1 Terabit-per-second (Tbps) signals over an equivalent distance of 3,200km. Each signal uses dual-polarisation 16-QAM (quadrature amplitude modulation) to achieve a 1.28Tbps signal. Thus each carrier holds 640Gbps: some 500Gbps data and the rest forward error correction (FEC) bits.
In current 100Gbps systems, dual-polarisation, quadrature phase-shift keying (DP-QPSK) modulation is used. Going from QPSK to 16-QAM doubles the bit rate. Bell Labs has also increased the symbol rate from some 30Gbaud to 80Gbaud using state-of-the-art high-speed electronics developed at Alcatel Thales III-V Lab.
"To achieve these rates, you need special high-speed components - multiplexers - and also high-speed multi-level devices," says Winzer. These are indium phosphide components, not CMOS and hence will not be deployed in commercial products for several years yet. "These things are realistic [in CMOS], just not for immediate product implementation," says Winzer.
Each carrier occupies 100GHz of channel bandwidth equating to 200GHz overall, or a 5.2b/s/Hz spectral efficiency. Current state-of-the-art 100Gbps systems use 50GHz channels, achieving 2b/s/Hz.
The 3,200km reach using 16-QAM technology is achieved in the lab, using good fibre and without any commercial product margins, says Winzer. Adding commercial product margins would reduce the optical link budget by 2-3dB and hence the overall reach.
Winzer says the one Terabit demonstration uses all the technologies employed in Alcatel-Lucent's photonic service engine (PSE) ASIC although the algorithms and soft-decision FEC used are more advanced, as expected in an R&D trial.
Before such one Terabit systems become commercial, progress in photonic integration will be needed as well as advances in CMOS process technology.
"Progress in photonic integration is needed to get opto-electronic costs down as it [one Terabit] is still going to need two-to-four sub-carriers," he says. A balance between parallelism and speed needs to be struck, and parallelism is best achieved using integration. "Without integration you are doubling up your expensive opto-electronic components which doesn't scale," says WInzer.
NeoPhotonics' PIC transceiver tackles PON business case
Gazettabyte completes its summary of optical announcements at ECOC, held in Amsterdam. In the third and final part, NeoPhotonics’ GPON multiport transceiver is detailed.
Part 3: NeoPhotonics
“Anything that can be done to get high utilisation of your equipment, which represents your up-front investment, helps the business case"
Chris Pfistner, NeoPhotonics
NeoPhotonics has announced a Gigabit passive optical network (GPON) transceiver designed to tackle the high up-front costs operators face when deploying optical access.
The GPON optical line terminal (OLT) transceiver has a split ratio of 1:128 - a passive optical network (PON) supporting 128 end points - yet matches the optical link budget associated with smaller split ratios. The transceiver, housed in an extended SFP module, has four fibre outputs, each supporting a conventional GPON OLT. The transceiver also uses a mode-coupling receiver implemented using optical integration.
According to NeoPhotonics, carriers struggle with the business case for PON given the relatively low take-up rates by subscribers, at least initially. “Anything that can be done to get high utilisation of your equipment, which represents your up-front investment, helps the business case,” says Chris Pfistner, vice president of product marketing at NeoPhotonics. “With a device like this, you can now cover four times the area you would normally cover.”
The GPON OLT transceiver, the first of a family, has been tested by operator BT that has described the technology as promising.
Reach and split ratio
The GPON transceiver supports up to 128 end points yet meets the GPON Class B+ 28dB link budget optical transceiver specification.
The optical link budget can be traded to either maximise the PON’s distance, limited due to the loss per fibre-km, or to support higher split ratios. However, a larger split ratio increases the insertion loss due to the extra optical splitter stages the signal passes through. Each 1:2 splitter introduces a 3.5dB loss, eroding the overall optical link budget and hence the PON’s reach.
GPON was specified with a Class B 20dB and Class C 30dB link budget. However once PON deployments started a 28dB Class B+ was created to match the practical requirements of operators. For Verizon, for example, a reach of 10-11km covers 95% of its single family units, says NeoPhotonics.
Operators wanting to increase the split ratio to 1:64 need an extra 4dB. This has led to the 32dB link budget Class C+. For shorter runs, in such cases as China, the Class C+ is used for a 1:128 split ratio. “They [operators] are willing to give up distance to cover an extra 1-by-2 split,” says Pfistner.
NeoPhotonics supports the 1:128 split ratio without suffering such loss by introducing two techniques: the mode-coupling receiver (MCR) and boosting the OLT transceiver's transmitter power.
A key issue dictating a PON performance is the sensitivity of the OLT's burst mode receiver. The upstream fibres are fed straight onto the NeoPhotonics’ MCR, eliminating the need for a 4x1 combiner (inverse splitter) and a resulting 6dB signal loss.
The GPON OLT transceiver showing the transmit and the mode-coupling receiver. Source: NeoPhotonics
The MCR is not a new concept, says Pfistner, and can be implemented straightforwardly using bulk optics. But such an implementation is relatively large. Instead, NeoPhotonics has implemented the MCR as a photonic integrated circuit (PIC) fitting the design within an extended SFP form factor.
“The PIC draws on our long experience of planar lightwave circuit technology, and [Santur’s] indium phosphide array technology, to do fairly sophisticated devices,” says Pfistner. NeoPhotonics acquired Santur in 2011.
The resulting GPON transceiver module fits within an SFP slot but it is some 1.5-2cm longer than a standard OLT SFP. Most PON line cards support four or eight OLT ports. Pfistner says a 1:4 ratio is the sweet spot for initial rollouts but higher ratios are possible.
On the transmit side, the distributed feedback (DFB) laser also goes through a 1:4 stage which introduces a 6dB loss. The laser transmit power is suitably boosted to counter the 6dB loss.
Operators
BT has trialled the optical performance of a transceiver prototype. “BT confirmed that the four outputs each represents a Class B+ GPON OLT output,” says Pfistner. Some half a dozen operators have expressed an interest in the transceiver, ranging from making a request to working with samples.
China is one market where such a design is less relevant at present. That is because China is encouraging through subsidies the rollout of PON OLTs even if the take-up rate is low. Pfistner, quoting an FTTH Council finding, says that there is a 5% penetration typically per year: “Verizon has been deploying PON for six years and has about a 30% penetration.”
Meanwhile, an operator only beginning PON deployments will first typically go after the neighbourhoods where a high take-up rate is likely and only then will it roll out PON in the remaining areas.
After five years, a 25% uptake is achieved, assuming this 5% uptake a year. At a 4x higher split ratio, that is the same bandwidth per user as a standard OLT in a quarter of the area, says NeoPhotonics.
“One big concern that we hear from operators is: Now I'm sharing the [PON OLT] bandwidth with 4x more users,” says Pfistner. “That is true if you believe you will get to the maximum number of users in a short period, but that is hardly ever the case.”
And although the 1:128 split ratio optical transceiver accounts for a small part of the carrier’s PON costs, the saving the MCR transceiver introduces is at the line card level. "That means at some point you are going to save shelves and racks [of equipment],” says Pfistner.
Roadmap
The next development is to introduce an MCR transceiver that meets the 32dB Class C+ specification. “A lot of carriers are about to make the switch from B+ to C+ in the GPON world,” says Pfistner. There will also be more work to reduce the size of the MCR PIC and hence the size of the overall pluggable form factor.
Beyond that, NeoPhotonics says a greater than 4-port split is possible to change the economics of 10 Gigabit PON, for GPON and Ethernet PON. “There are no deployments right now because the economics are not there,” he adds.
“The standards effort is focussed on the 'Olympic thought': higher bandwidth, faster, further reach, mode-coupling receiver (MCR) whereas the carriers focus is: How do I lower the up-front investment to enter the FTTH market?” says Pfistner.
Further reading:
GPON SFP Transceiver with PIC based Mode-Coupled Receiver, Derek Nesset, David Piehler, Kristan Farrow, Neil Parkin, ECOC Technical Digest 2012 paper.
Lightwave: Mode coupling receiver increases PON split ratios, click here
Ovum: Lowering optical transmission cost at ECOC 2012, click here
Summary Gazettabyte stories from ECOC 2012, click here
60-second interview with .... Sterling Perrin
Heavy Reading has published a report Photonic Integration, Super Channels & the March to Terabit Networks. In this 60-second interview, Sterling Perrin, senior analyst at the market research company, talks about the report's findings and the technology's importance for telecom and datacom.
"PICs will be an important part of an ensemble cast, but will not have the starring role. Some may dismiss PICs for this reason, but that would be a mistake – we still need them."
Sterling Perrin, Heavy Reading
Heavy Reading's previous report on optical integration was published in 2008. What has changed?
The biggest change has been the rise of coherent detection, bringing electronics to prominence in the world of optics. This is a big shift - and it has taken some of the burden off photonic integration. Simply put, electronics has taken some of the job away from optics.
How important is optical Integration, for optical component players and for system vendors?
Until now, photonic integration has not been a ‘must have’ item for systems suppliers. For the most part, there have been other ways to get at lower costs and footprint reductions.
I think we are starting to see photonic integration move into the must-have category for systems suppliers, in certain applications, which means that it becomes a must-have item for the components companies that supply them.
How should one view silicon photonics and what importance does Heavy Reading attach to Cisco System's acquisition of silicon photonics' startup, Lightwire?
When we published the last [2008] report, silicon photonics was definitely within the hype cycle. We’ve seen the hype fade quite a bit – it’s now understood that just because a component is made with silicon, it’s not automatically going to be cheaper. Also, few in the industry continue to talk about a Moore’s Law for optics today. That said, there are applications for silicon photonics, particularly in data centre and short-reach applications, and the technology has moved forward.
Cisco’s acquisition of Lightwire is a good testament for how far the technology has come. This is a strategic acquisition, aimed at long-term differentiation, and Cisco believes that silicon photonics will help them get there.
"It will be interesting to watch what other [optical integration] M&A activity occurs, and how this activity affects the components players"
What are the main optical integration market opportunities?
In long haul, we already see applications for photonic integrated circuits (PICs). Certainly, Infinera’s PIC-based DTN and DTN-X systems stand out. But also, the OIF has specified photonic integration in its 100 Gigabit long haul, DWDM (dense wavelength division multiplexing) MSA (multi-source agreement) – it was needed to get the necessary size reduction.
Moving forward, there is opportunity for PICs in client-side modules as PICs are the best way to reduce module sizes and improve system density. Then, beyond 100G, to super-channel-based long-haul systems, PICs will play a big role here, as parallel photonic integration will be used to build these super-channels.
Were you surprised by any of the report's findings?
When I start researching a report, I am always hopefully for big black and white kinds of findings – this is the biggest thing for the industry or this is a dud. With photonic integration, we found such a wide array of opinions and viewpoints that, in the end, we had to place photonic integration somewhere in the middle.
It’s clear that system vendors are going to need PICs but it’s also clear that PICs alone won’t solve all the industry’s challenges. PICs will be an important part of an ensemble cast, but will not have the starring role. Some may dismiss PICs for this reason, but that would be a mistake – we still need them.
What optical integration trends/ developments should be watched over the next two years?
The year started with two major system suppliers buying PIC companies: Cisco and Lightwire and Huawei and CIP Technologies. With Alcatel-Lucent having in-house abilities, and, of course, Infinera, this should put pressure on other optical suppliers to have a PIC strategy.
It will be interesting to watch what other M&A activity occurs, and how this activity affects the components players.
The editor of Gazettabyte worked with Heavy Reading in researching photonic integration for the report.
OneChip Photonics targets the data centre with its PICs
OneChip Photonics is developing integrated optical components for the IEEE 40GBASE-LR4 and 100GBASE-LR4 interface standards.
The company believes its photonic integrated circuits (PICs) will more than halve the cost of the 40 and 100 Gigabit 10km-reach interfaces, enough for LR4 to cost-competitively address shorter reach applications in the data centre.
"I think we can cut the price [of LR4 modules] by half or better”
Andy Weirich, OneChip Photonics
The products mark an expansion of the Canadian startup's offerings. Until now OneChip has concentrated on bringing PIC-based passive optical network (PON) transceivers to market.
LR4 PICs
The startup is developing separate LR4 transmitter and receiver PICs. The 40 and 100GBASE-LR4 receivers are due in the third quarter of 2012, while the transmitters are expected by the year end.
The 40GBASE-LR4 receiver comprises a wavelength demultiplexer - a 4-channel arrayed waveguide grating (AWG) - and four photo-detectors operating around 1300nm. A spot-size converter - an integrated lens - couples the receiver's waveguide's mode field to the connecting fibre.
"[Data centre operators] are saying that they are having to significantly bend out of shape their data centre architecture to accommodate even 300m reaches”
The 40GBASE-LR4 transmitter PIC comprises four directly-modulated distributed feedback (DFB) lasers while the 100GBASE-LR4 use four electro-absorption modulator DFB lasers. Different lasers for the two PICs are required since the four wavelengths at 100 Gig, also around 1300nm, are more tightly spaced: 5nm versus 20nm. "They are much closer together than the 40 Gig version,” says Andy Weirich, OneChip Photonics' vice president of product line management.
Another consequence of the wider wavelength spacings is that the 40 Gig transmitter uses four discrete lasers. “Because the 40 Gig wavelengths are much further apart, putting all the lasers on the one die is problematic," says Weirich. The 40GBASE-LR4 design thus uses five indium phosphide components: four lasers and the AWG, while the 40GBASE-LR4 receiver and the two 100GBASE-LR4 devices are all monolithic PICs.
Both LR4 transmitter designs also include monitor photo-diodes for laser control
Lower size and cost
OneChip says the resulting PICs are tiny, measuring less than 3mm in length. “We think the PICs will enable the packaging of LR4 in a QSFP,” says Weirich. 40GBASE-LR4 products already exists in the QSFP form factor but the 100GBASE-LR4 uses a CFP module.
The startup expects module makers to use its receiver chips once they become available rather than wait for the receiver-transmitter PIC pair. "Reducing the size of one half the solution is possibly good enough to fit the whole hybrid design - the PIC for the receive and discretes for the transmit - into a QSFP,” says Weirich.
The PICs are expected to reduce significantly the cost of LR4 modules. "I think we can cut the price by half or better,” says Weirich. “Right now the LR4 is far too expensive to be used for data centre interconnect.” OneChip expects its LR4 PICs to be cost-competitive with the 2km reach 10x10 MSA interface.
Meanwhile, short-reach 40 and 100 Gig interfaces use VCSEL technology and multi-mode fibre to address 100m reach requirements. In larger data centres this reach is limiting. Extended reach - 300-400m - multimode interfaces have emerged but so far these are at 40 Gig only.
"[Data centre operators] are saying that they are having to significantly bend out of shape their data centre architecture to accommodate even 300m reaches,” says Weirich. “They really want more than that.”
OneChip believes interfaces distances of 200m-2km is underserved and it is this market opportunity that it is seeking to address with its LR4 designs.
Roadmap
Will OneChip integrate the design further to product a single PIC LR4 transceiver?
"It can be put into one chip but it is not clear that there is an economic advantage,” says Weirich. Indeed one PIC might even be more costly than the two-PIC chipset.
Another factor is that at 100 Gig, the 25Gbps electronics present a considerable signal integrity design challenge. “It is very important to keep the electronics very close to the photo-detectors and the modulators,” he says. “That becomes more difficult if you put it all on the one chip.” The fabrication yield of a larger single PIC would also be reduced, impacting cost.
OneChip, meanwhile, has started limited production of its PON optical network unit (ONU) transceivers based on its EPON and GPON PICs. The company's EPON transceivers are becoming generally available while the GPON transceivers are due in two months’ time.
The company has yet to decide whether it will make its own LR4 optical modules. For now OneChip is solely an LR4 component supplier.
Further reading:
See OFC/ NFOEC 2012 highlights, the Kotura story in the Optical Engines section
NeoPhotonics secures PIC specialist Santur
NeoPhotonics has completed the acquisition of Santur, the tunable laser and photonic integration specialist, boosting the company's annual turnover to a quarter of a billion dollars.
Source: Gazettabyte
The acquisition helps NeoPhotonics become a stronger, vertically integrated transponder supplier. In particular, it broadens NeoPhotonics’ 40 and 100 Gigabit-per-second (Gbps) component portfolio, turns the company into a leading provider of tunable lasers and enhances its photonic integration expertise.
“Our business over a number of years has grown as the importance of photonic integrated circuits and the products deriving from them have grown,” says Tim Jenks, CEO of NeoPhotonics. “We believe it is a critical part of the network architecture today and going forward.”
Some US $39.2M in cash has been paid for Santur, and could be up to $7.5M more depending on Santur’s products' market performance over the next year.
NeoPhotonics has largely focussed on telecom but Jenks admits it is broadening its offerings. “Certainly a very significant portion of fibre-optic components are consumed in data and storage, and while historically that has not been a significant part of NeoPhotonics, it is a large and important market overall,” says Jenks.
"It [optical components] portends the future of the technology industry"
Tim Jenks, NeoPhotonics
The company will continue to address telecom but will add products to additional segments, including datacom. In July, the company announced its first CFP module supporting the 40 Gigabit Ethernet (GbE) 40GBASE-LR4 standard. Santur also supplies 40Gbps and 100Gbps 10km transceivers, in QSFP and CFP form factors, respectively.
Santur made its name as a tunable laser supplier and is estimated to have a 50% market share, according to Ovum. More recently it has developed arrays of 10Gbps transmitters. Such photonic integrated circuits (PICs) are used for the 10x10 multi-source agreement (MSA).
The acquisition complements NeoPhotonics’ 40Gbps and 100Gbps integrated indium-phosphide receiver components, enabling the company to provide the various optical components needed for 40 and 100Gbps modules. Santur also has narrow line-width tunable laser technology used at the coherent transmitter and receiver. But Jenks confirms that the company has not announced a transmitter at 28Gbps using this narrow line-width laser.
10x10 MSA
Santur has been a key player in the 10x10 MSA, developed as a low cost competitor to the IEEE 100 Gigabit Ethernet (GbE) 10km 100GBASE-LR4 and 40km -ER4 standards.
Large content service providers such as Google want cheaper 100GbE interfaces and the 10x10 MSA module, built using 10x10Gbps electrical and optical interfaces, is approximately half the cost of the IEEE interfaces.
"There is an opportunity with the 10x10 MSA," says Jenks. "The 10x10 does not require the gearbox IC, it is therefore lower cost and lower power, and fulfills a need that a 4x25Gig, with a rather immature technology and a requirement for a gearbox IC, does not."
In August the 10x10 MSA announced further specifications: a 10km version of the 10x10 MSA as well as two 40km-reach WDM interfaces: a 4x10x10Gbps and an 8x10x10Gbps. "There are end users that want to use these," says Jenks.
“The ability for a system vendor to lead is a challenging task. For a system vendor to lead and simultaneous lead in developing their componentry is a daunting task.”
Acquisitions
NeoPhotonics has made several acquisitions over the years, including four in 2006 (see chart). But Santur's revenues - some $50m - are larger than the aggregated revenues of all the previous acquisitions.
"I think of acquisitions as being inorganic for maybe two years and after that they are all organic," says Jenks. The acquisitions have helped NeoPhotonics broaden its technologies, strengthen the company's know-how and acquire customers and relationships.
“If someone says what did you do with this product from that company, they are asking the wrong question,” says Jenks. “By the law of averages, some [acquisitions] do better, some do worse but overall it has been quite successful.”
System vendors and vertical integration
Jenks says he is aware of system vendors taking steps to develop components and technology in-house but he does not believe this will change the primary role of the component vendors.
"Equipment vendors are building some things in-house for a near-term cost advantage, better insights into cost of production or better insights in how the technology can go,” says Jenks. “All reasons to have some form of vertical integration.”
But in technology leadership, no one company has a monopoly of talent. As such vertical integration is a double-edged sword, he says, a company can become quite expert but it can also isolate itself from what the rest of the world is doing.
“The ability for a system vendor to lead is a challenging task,” says Jenks. “For a system vendor to lead, and simultaneous lead in developing their componentry, is a daunting task.”
The world is flat
Jenks, whose background is in mechanical and nuclear engineering, highlights two aspects that strike him about the optical component industry.
One is that telecoms is ubiquitous and because optical components go into telecoms, optical components is a global industry. "The world is very flat in optical components,” he says.
Second, the hurdles to undertake experiments in optical components is lower than the significant capital investment needed for nuclear engineering, for example. "Colleges and universities turn out graduates in physics and electrical engineering that are well trained and need a lighter physical plant,” says Jenks. This aspect of the education promotes a globally diverse and a rather 'flat' industry.
“When I go to a trade show in China, Europe or the US, I'm running into colleagues from the industry that I know from each country we do business, and that is a lot of countries,” he says.
All this, for Jenks, makes optical components a fascinating industry, one that is on the leading edge of technology and also industrial trend.
"It [optical components] portends the future of the technology industry: flatter and flatter with more global players and more global competition," says Jenks. “At the moment it is novel in optical components but in a few years' time it won't be unique to optical components.”
NeoPhotonics at a glance
The company segments its revenues into the areas of speed and agility (10-100Gbps products, planar lightwave circuits - ROADMs, arrayed waveguide gratings), access (FTTh, cable TV, wireless backhaul) and SDH and slow-speed DWDM, products designed 3-5 years ago.
Historically these three segments' revenues have been equal but this year the access business has been larger, accounting for 40% of revenues due to China's huge FTTx rollout.
Huawei is NeoPhotonics' largest customer. “They have been as much as half our revenue," says Jenks. And depending on the quarter, Ciena and Alcatel-Lucent have been reported as 10% customers.
High fives: 5 Terabit OTN switching and 500 Gig super-channels.
Infinera has announced a core network platform that combines Optical Transport Network (OTN) switching with dense wavelength division multiplexing (DWDM) transport. "We are looking at a system that integrates two layers of the network," says Mike Capuano, vice president of corporate marketing at Infinera.
"This is 100Tbps of non-blocking switching, all functioning as one system. You just can't do that with merchant silicon."
Mike Capuano, Infinera
The DTN-X platform is based on Infinera's third-generation photonic integrated circuit (PIC) that supports five, 100Gbps coherent channels.
Each DTN-X platform can deliver 5 Terabits-per-second (Tbps) of non-blocking OTN switching using an Infinera-designed ASIC. Ten DTN-X platforms can be combined to scale the OTN switching and transport capacity to 50Tbps currently.
Infinera also plans to add Multiprotocol Label Switching (MPLS) to turn the DTN-X into a hybrid OTN/ MPLS switch. With the next upgrades to the PIC and the switching, the ten DTN-X platforms will scale to 100Tbps optical transport and 100Tbps OTN and MPLS switching capacity.
The platform is being promoted by Infinera as a way for operators to tackle network traffic growth and support developments such as cloud computing where applications and content increasingly reside in the network. "What that means [for cloud-based services to work] is a network with huge capacity and very low latency," says Capuano.
Platform details
The 5x100Gbps PIC supports what Infinera calls a 500Gbps 'super-channel'. Each super-channel is a multi-carrier implementation comprising five, 100Gbps wavelengths. Combined with OTN, the 500Gbps super-channel can be filled with 1, 10, 40 and 100 Gigabit streams (SONET/SDH, Ethernet, video etc). Moreover, there is no spectral efficiency penalty: the super-channel uses 250GHz of fibre spectrum, provisioning five 50GHz-wide, 100Gbps wavelengths at a time.
"We have seen 40 and 100Gbps come on the market and they are definitely helping with fibre capacity issues," says Capuano. "But they are more expensive from a cost-per-bit perspective than 10Gbps." By introducing the 500Gbps PIC, Infinera says it is reducing the cost-per-bit performance of high speed optical transport.
DTN-X: shown are 5 line and tributary cards top and bottom with switching cards in the centre of the chassis. Source: Infinera
Integrating OTN switching within the platform results in the lowest cost solution and is more efficient when compared to multiplexed transponders (muxponder) configured manually, or an external OTN switch which must be optically connected to the transport platform.
The DTN-X also employs Generalised MPLS (GMPS) software. "GMPLS makes it easy to deploy networks and services with point-and-click provisioning," says Capuano.
Each DTX-N line card supports a 500Gbps PIC but the chassis backplane is specified at 1Tbps, ready for Infinera's next-generation 10x100Gbps PIC that will upgrade the DTN-X to a 10Tbps system. "We have already presented our test results for our 1Tbps PIC back in March," says Capuano. The fourth-generation PIC, estimated around 2014 (based on a company slide although Infinera has made no public comment), will support a 1Tbps super-channel.
Adding MPLS will add the transport capability of the protocol to the DTN-X. "You will have MPLS transport, OTN switching and DWDM all in one platform," says Capuano.
OTN switching is the priority of the tier-one operators to carry and process their SONET/SDH traffic; adding MPLS will enable extra traffic processing capabilities to the system, he says.
Infinera says that by eventually integrating MPLS switching into the optical transport network, operators will be able to bypass expensive router ports and simplify their network operation.
Performance
Infinera says that the DTX-N 5Tbps performance does not dip however the system is configured: whether solely as a switch (all line card slots filled with tributary modules), mixed DWDM/ switching (half DWDM/ half tributaries, for example) or solely as a DWDM platform. Depending on the cards in the DTN-X platform, the transport/ switching configuration can be varied but the 5Tbps I/O capacity is retained. Infinera says other switches on the market do lose I/O capacity as the interface mix is varied.
Overall, Infinera claims the platform requires half the power of competing solutions and takes up a third less space.
The DTN-X will be available in the first half of 2012.
Analysis
Gazettabyte asked several market research firms about the significance of the DTN-X announcement and the importance of combining OTN, DWDM and soon MPLS within one platform.
Ovum
Ron Kline, principal analyst, and Dana Cooperson, vice president, of the network infrastructure practice
"MPLS switching is setting up a very interesting competitive dynamic among vendors"
Dana Cooperson, Ovum
The DTN-X is a platform for the largest service providers and their largest sites, says Ovum.
It sees the DTN-X in the same light as other integrated OTN/ WDM platforms such as Huawei's OSN 8800, Nokia Siemens Networks' hiT 7100, Alcatel-Lucent's 1830 PSS and Tellabs' 7100 OTS.
"It fits the mold for Verizon's long-haul optical transport platform (LH OTP), especially once MPLS is added," says Kline. "NSN is also claiming it will add MPLS to the 7100. Once MPLS is added, then you have the big packet optical transport box that Verizon wants."
The DTN-X platform will boost the business case for 100 Gig in a similar way to how Infinera's current PIC has done at 10 Gig. "The others will be forced to lower price," says Kline.
Having GMPLS is important, especially if there is a need to do dynamic bandwidth allocation, however it is customer-dependent. "When you start digging, it's hard to find large-scale implementations of GMPLS," says Kline.
The Ovum analysts stress that the need for OTN in the core depends on the customer. Content service providers like Google couldn't care less about OTN. "It's really an issue for multi-service providers like BT and AT&T," says Cooperson,
There is a consensus about the need for MPLS in the core. "Different service providers are likely to take different approaches — some might prefer an integrated box and others might not, it depends on their business," she says. "I think MPLS switching is setting up a very interesting competitive dynamic among vendors that focus on IP/MPLS, those that focus on optical, and those that are trying to do both [optical and IP/MPLS].
Ovum highlights several aspects regarding the DTN-X's claimed performance.
"Assuming it performs as advertised, this should finally give Infinera what it needs to be of real interest to the tier-ones," says Cooperson. "The message of scalability, simplicity, efficiency, and profitability is just what service providers want to hear."
Cooperson also highlights Infinera's approach to optical-electrical-optical conversion and the benefit this could deliver at line speeds greater than 100Gbps.
At present ROADMs are being upgraded to support flexible spectrum channel configurations, also known as gridless. This is to enable future line speeds that will use more spectrum than current 50GHz DWDM channels. Operators want ROADMs that support flexible spectrum requirements but managing the network to support these variable width channels is still to resolved.
"It fits the mold for Verizon's long-haul optical transport platform (LH OTP), especially once MPLS is added"
Ron Kline, Ovum
Infinera's approach is based on conversion to the electrical domain when dropping and regenerating wavelengths such that the issue of flexible channels does not arise or is at least forestalled. This, says Cooperson, could be Infinera's biggest point of differentiation.
"What impresses me is the 500Gbps super-channel using five, 100Gbps carriers and the size of the switch fabric," adds Kline. The 5Tbps switching performance also exceeds that of everyone else: "Alcatel-Lucent is closest with 4Tbps but most range from 1-3Tbps and top out at 3Tbps."
The ease of use is also a big deal. Infinera did very well in marketing rapid turn up: 10 Gig in 10 days for example, says Kline: "It looks like they will be able to do the same here with 100 Gig."
Infonetics Research
Andrew Schmitt, directing analyst, optical
"GMPLS isn't that important, yet."
The DTN-X is a WDM platform which optionally includes a switch fabric for carriers that want it integrated with the transport equipment, says Schmitt. Once MPLS is added, it has the potential to be a full-blown packet-optical system.
"[The announcement is] pretty significant though not unexpected," says Schmitt. "I think the key question is what it costs, and whether the 500G PIC translates into compelling savings."
Having MPLS support is important for some carriers such as XO Communications and Google but not for others.
Schmitt also says GMPLS isn't that important, yet. "Infinera's implementation of regen-rich networks should make their GMPLS implementation workable," he says. "It has been building networks like that for a while."
OTN in the core is still an open debate but any carrier that doesn't have the luxury of a homogenous data network needs it, he says
Schmitt has yet to speak with carriers who have used the DTN-X: "I can't comment on claimed performance but like I said, cost is important."
ACG Research
Eve Griliches, managing partner
"Infinera has already introduced the 500G PIC, but the OTN is significant in that it can be used as a standalone OTN switch, and it has the largest capacity out there today"
The DTN-X as an OTN/ WDM platform awaiting label switch router (LSR) functionality, says Griliches: "With the LSR functionality it will be able to do statistical multiplexing for direct router connections."
Infinera has already introduced the 500 Gig PIC but the OTN is significant in that it can be used as a standalone OTN switch, and it has the largest capacity out there today. An OTN survey conducted last year by ACG Research found that the switch capacity sweet spot is between 4 and 8Tbps.
Griliches says that LSR-based products are taking time to incorporate WDM and OTN technologies, while it is unclear when the DTN-X will support MPLS to add LSR capabilities. The race is on as to whom can integrate everything first, but DWDM and OTN before MPLS is the right direction for most tier-one operators, she says.
Infinera has over eight thousand of its existing DTNs deployed at 85 customers in 50 countries. The scale of the DTN-X will likely broaden Infinera's customer base to include tier-one operators, says Griliches.
ACG Research has heard positive feedback from operators it has spoken to. One stressed that the decreased port count due to the larger OTN cross-connect significantly improves efficiencies. Another operator said it would pick Infinera and said the beta version of the 500Gbps PIC is "working beautifully".
Infinera details Terabit PICs, 5x100G devices set for 2012
Infinera has given first detail of its terabit coherent detection photonic integrated circuits (PICs). The pair - a transmitter and a receiver PIC – implement a ten-channel 100 Gigabit-per-second (Gbps) link using polarisation multiplexing quadrature phase-shift keying (PM-QPSK). The Infinera development work was detailed at OFC/NFOEC held in Los Angeles between March 6-10.
Infinera has recently demonstrated its 5x100Gbps PIC carrying traffic between Amsterdam and London within Interoute Communications’ pan-European network. The 5x100Gbps PIC-based system will be available commercially in 2012.
“We think we can drive the system from where it is today – 8 Terabits-per-fibre - to around 25 Terabits-per-fibre”
Dave Welch, Infinera
Why is this significant?
The widespread adoption of 100Gbps optical transport technology will be driven by how quickly its cost can be reduced to compete with existing 40Gbps and 10Gbps technologies.
Whereas the industry is developing 100Gbps line cards and optical modules, Infinera has demonstrated a 5x100Gbps coherent PIC based on 50GHz channel spacing while its terabit PICs are in the lab.
If Infinera meets its manufacturing plans, it will have a compelling 100Gbps offering as it takes on established 100Gbps players such as Ciena. Infinera has been late in the 40Gbps market, competing with its 10x10Gbps PIC technology instead.
40 and 100 Gigabit
Infinera views 40Gbps and 100Gbps optical transport in terms of the dynamics of the high-capacity fibre market. In particular what is the right technology to get most capacity out of a fibre and what is the best dollar-per-Gigabit technology at a given moment.
For the long-haul market, Dave Welch, chief strategy officer at Infinera, says 100Gbps provides 8 Terabits (Tb) of capacity using 80 channels versus 3.2Tb using 40Gbps (80x40Gbps). The 40Gbps total capacity can be doubled to 6.4Tb (160x40Gbps) if 25GHz-spaced channels are used, which is Infinera’s approach.
“The economics of 100 Gigabit appear to be able to drive the dollar-per-gigabit down faster than 40 Gigabit technology,” says Welch. If operators need additional capacity now, they will adopt 40Gbps, he says, but if they have spare capacity and can wait till 2012 they can use 100Gbps. “The belief is that they [operators] will get more capacity out of their fibre and at least the same if not better economics per gigabit [using 100Gbps],” says Welch. Indeed Welch argues that by 2012, 100Gbps economics will be superior to 40Gbps coherent leading to its “rapid adoption”.
For metro applications, achieving terabits of capacity in fibre is less of a concern. What matters is matching speeds with services while achieving the lowest dollar-per-gigabit. And it is here – for sub-1000km networks – where 40Gbps technology is being mostly deployed. “Not for the benefit of maximum fibre capacity but to protect against service interfaces,” says Welch, who adds that 40 Gigabit Ethernet (GbE) rather than 100GbE is the preferred interface within data centres.
Shorter-reach 100Gbps
Companies such as ADVA Optical Networking and chip company MultiPhy highlight the merits of an additional 100Gbps technology to coherent based on direct detection modulation for metro applications (for a MultiPhy webinar on 100Gbps direct detection, click here). Direct detection is suited to distances from 80km up to 1000km, to connect data centres for example.
Is this market of interest to Infinera? “This is a great opportunity for us,” says Welch.
The company’s existing 10x10Gbps PIC can address this segment in that it is least 4x cheaper than emerging 100Gbps coherent solutions over the next 18 months, says Welch, who claims that the company’s 10x10Gbps PIC is making ‘great headway’ in the metro.
“If the market is not trying to get the maximum capacity but best dollar-per-gigabit, it is not clear that full coherent, at least in discrete form, is the right answer,” says Welch. But the cost reduction delivered by coherent PIC technology does makes it more competitive for cost-sensitive markets like metro.
A 100Gbps coherent discrete design is relatively costly since it requires two lasers (one as a local oscillator (LO - see fig 1 - at the receiver), sophisticated optics and a high power-consuming digital signal processor (DSP). “Once you go to photonic integration the extra lasers and extra optics, while a significant engineering task, are not inhibitors in terms of the optics’ cost.”
Coherent PICs can be used ‘deeper in the network’ (closer to the edge) while shifting the trade-offs between coherent and on-off keying. However even if the advent of a PIC makes coherent more economical, the DSP’s power dissipation remains a factor regarding the tradeoff at 100Gbps line rates between on-off keying and coherent.
Welch does not dismiss the idea of Infinera developing a metro-centric PIC to reduce costs further. He points out that while such a solution may be of particular interest to internet content companies, their networks are relatively simple point-to-point ones. As such their needs differ greatly from cable operators and telcos, in terms of the services carried and traffic routing.
PIC challenges
Figure 1: Infinera's terabit PM-QPSK coherent receiver PIC architecture
There are several challenges when developing multi-channel 100Gbps PICs. “The most difficult thing going to a coherent technology is you are now dealing with optical phase,” says Welch. This requires highly accurate control of the PIC’s optical path lengths.
The laser wavelength is 1.5 micron and with the PIC's indium phosphide waveguides this is reduced by a third to 0.5 micron. Fine control of the optical path lengths is thus required to tenths of a wavelength or tens of nanometers (nm).
Achieving a high manufacturing yield of such complex PICs is another challenge. The terabit receiver PIC detailed in the OFC paper integrates 150 optical components, while the 5x100Gbps transmit and receive PIC pair integrate the equivalent of 600 optical components.
Moving from a five-channel (500Gbps) to a ten-channel (terabit) PIC is also a challenge. There are unwanted interactions in terms of the optics and the electronics. “If I turn one laser on adjacent to another laser it has a distortion, while the light going through the waveguides has potential for polarisation scattering,” says Welch. “It is very hard.”
But what the PICs shows, he says, is that Infinera’s manufacturing process is like a silicon fab’s. “We know what is predictable and the [engineering] guys can design to that,” says Welch. “Once you have got that design capability, you can envision we are going to do 500Gbps, a terabit, two terabits, four terabits – you can keep on marching as far as the gigabits-per-unit [device] can be accomplished by this technology.”
The OFC post-deadline paper details Infinera's 10-channel transmitter PIC which operates at 10x112Gbps or 1.12Tbps.
Power dissipation
The optical PIC is not what dictates overall bandwidth achievable but rather the total power dissipation of the DSPs on a line card. This is determined by the CMOS process used to make the DSP ASICs, whether 65nm, 40nm or potentially 28nm.
Infinera has not said what CMOS process it is using. What Infinera has chosen is a compromise between “being aggressive in the industry and what is achievable”, says Welch. Yet Infinera also claims that its coherent solution consumes less power than existing 100Gbps coherent designs, partly because the company has implemented the DSP in a more advanced CMOS node than what is currently being deployed. This suggests that Infinera is using a 40nm process for its coherent receiver ASICs. And power consumption is a key reason why Infinera is entering the market with a 5x100Gbps PIC line card. For the terabit PIC, Infinera will need to move its ASICs to the next-generation process node, he says.
Having an integrated design saves power in terms of the speeds that Infinera runs its serdes (serialiser/ deserialiser) circuitry and the interfaces between blocks. “For someone else to accumulate 500Gbps of bandwdith and get it to a switch, this needs to go over feet of copper cable, and over a backplane when one 100Gbps line card talks to a second one,” says Welch. “That takes power - we don’t; it is all right there within inches of each other.”
Infinera can also trade analogue-to-digital (A/D) sampling speed of its ASIC with wavelength count depending on the capacity required. “Now you have a PIC with a bank of lasers, and FlexCoherent allows me to turn a knob in software so I can go up in spectral efficiency,” he says, trading optical reach with capacity. FlexCoherent is Infinera’s technology that will allow operators to choose what coherent optical modulation format to use on particular routes. The modulation formats supported are polarisation multiplexed binary phase-shift keying (PM-BPSK) and PM-QPSK.
Dual polarisation 25Gbaud constellation diagrams
What next?
Infinera says it is an adherent of higher quadrature amplitude modulation (QAM) rates to increase the data rate per channel beyond 100Gbps. As a result FlexCoherent in future will enable the selection of higher-speed modulation schemes such as 8-QAM and 16-QAM. “We think we can drive the system from where it is today –8 Terabits-per-fibre - to around 25 Terabits-per-fiber.”
But Welch stresses that at 16-QAM and even higher level speeds must be traded with optical reach. Fibre is different to radio, he says. Whereas radio uses higher QAM rates, it compensates by increasing the launch power. In contrast there is a limit with fibre. “The nonlinearity of the fibre inhibits higher and higher optical power,” says Welch. “The network will have to figure out how to accommodate that, although there is still significant value in getting to that [25Tbps per fibre]” he says.
The company has said that its 500 Gigabit PIC will move to volume manufacturing in 2012. Infinera is also validating the system platform that will use the PIC and has said that it has a five terabit switching capacity.
Infinera is also offering a 40Gbps coherent (non-PIC-based) design this year. “We are working with third-party support to make a module that will have unique performance for Infinera,” says Welch.
The next challenge is getting the terabit PIC onto the line card. Based on the gap between previous OFC papers to volume manufacturing, the 10x100Gbps PIC can be expected in volume by 2014 if all goes to plan.