Paying homage to Harald Bock
Harald Bock, described by an optical networking executive as one of the great people of our industry, has died. Former colleagues describe the man and their sense of loss
Those who knew and worked with Harald Bock have been stunned by his sudden passing at 55. For them, Harald was a valued and much-admired friend, a deep thinker who made his views heard, quietly yet powerfully.
Last February, Harald changed jobs, becoming chief product officer at Ekinops after six years at Infinera. He was excited by the role and enjoyed his introductory period travelling to Ekinops’ sites, meeting colleagues and customers, and working on the company’s strategy.
Sylvain Quartier, Ekinops’ chief marketing & strategy officer, says it took the company a year to find the right candidate. Ekinops knew of Harald’s optical networking expertise but was impressed with his keenness to expand into what, for him, were new product areas such as routeing and cybersecurity.
“We needed someone expert in one domain and with good experience in product strategy,” says Quartier. “He was full of joy and happy to work.”
During Harald’s short spell at the company, he sharpened Ekinops’s product plans. “We’re executing his roadmap and strategy today,” says Quartier. “In six months, he had a great impact.”
Career
Harald earned his PhD in physics, specialising in polymer materials.
“Polymers may become an important material system for future high-speed [optical] modulators,” says Uwe Fischer, who was chief technology officer (CTO) at former optical networking firm, Coriant. “He was ahead of his time by doing something in his PhD thesis which is about to become important in business and technology.”
Harald’s career spanned some notable optical networking firms: Marconi, Nokia Siemens Networks, Coriant, and Infinera. He was part of Uwe Fischer’s team at Nokia Siemens Networks and Coriant. Harald’s strength was as a technologist, and had roles in several CTO offices.
Stefan Voll, then a lead product line manager and now senior director of business development at Adtran, worked with Harald at Nokia Siemens Networks in 2012.
The two were tasked with carving out the optical business of Nokia Siemens Networks in what was to become Coriant. “The carve-out was a big achievement,” says Voll. Harald represented the CTO office and Voll led the product line manager team and the two were tasked with making the product portfolio not only viable but profitable. This required aligning technical aspects with business needs, setting the foundation for Coriant’s operations.
At Coriant, Harald contributed to the development of Groove, one of the first compact modular platform for metro wavelength division multiplexing (WDM) and long-haul networks. Harald continued in the CTO office at Infinera after it acquired Coriant in 2019.
More recently, Harald served as editor of the Optical Internetworking Forum’s (OIF) 1600ZR project, taking over the role after an OIF member stepped down. The work involves standardising 1600-gigabit coherent optics for high-capacity networks. “He stepped in as smooth as possible,” says Karl Gass, optical vice chair of the OIF’s physical link layer (PLL) working group. “He knew how to manage personalities and get things done.”
Work ethic
Christian Uremovic, senior director of solution marketing at Nokia Optical Networks, worked with Harald at Coriant and Infinera. “He was a respected and valued technology guide for product line management and sales and for us in marketing,” says Uremovic. It wasn’t always an easy role; groups in the company would pull in different directions, and bringing it all together was a challenge. “Sometimes you had to make difficult decisions,” says Uremovic.
“He would like to understand basic technologies, and when, at executive meetings, he wanted to bring his opinion and convince others, he would talk quietly,” says Ekinop’s Quartier. “Everybody would be quiet because you wanted to hear him, and he made a strong impression.” With this quiet manner, Harald would progress the discussion and bring everybody in the right direction, says Quartier.
It is something Robert Richter, managing director and senior vice president, customer executive, product marketing office, at Nokia, highlights: “Harald would reiterate his view calmly, even if it annoyed some leaders, but it was always positive,” he says, adding that Harald was not the kind of guy who did what management told him. But he was always trying to change opinions constructively.
Voll reinforces this, saying Harald would not let shortcuts slide, ensuring all critical aspects were covered. “He was not afraid of conflict.” Harald would broaden discussions to bring in new angles to the point where it could be annoying. “But it was always valuable,” says Voll. “He was not fast, and that’s because he was a deep thinker. He reacted in meetings, but not immediately; he needed some time putting his thoughts together.”
Voll says Harald was focused on long-term technology trends, five to six years out, compared to product line managers’ shorter-term view. “He taught me to assess competitiveness through metrics like capacity per power consumption,” says Voll. Harald would say: “Convert it into capacity per volume or per power consumption.” Ten years ago, these were not usual metrics, but Harald used them to measure product plans. Harald would also also look product optimisations, such as whether a platform’s chips had features that were not used.
Maxim Kuschnerov, Director of R&D, at Huawei, worked with Harald at Coriant. He recalls a meeting he had with Deutsche Telekom presenting Huawei’s research topics. Afterwards, a Deutsche Telekom executive remarked that although it was framed as a research discussion, Kuschnerov kept steering the conversation toward commercial applications and customer value: “It reminds me of a guy at Coriant who talks about use cases a lot — Harald Bock”. Smiling, Kuschnerov replied: “Where do you think I was taught to think like that? I worked with Harald!”
Traits
Ekinops’ Quartier highlights Harald’s general cheerfulness: “He was always smiling, which was much appreciated.”
Richter, who worked with Harald for 18 years at Nokia Siemens Networks and Coriant, describes him as the most empathetic person he has met. “He was always listening to people and was very patient,” he says. Harald would bring cakes to the office from vacations and would never speak ill of others. He’d keep positive even during challenging times and in that sense, he was a role model, says Richter: “It was always a pleasure to talk to him over a coffee. He was open to a dialogue.”
Harald had an impressive fitness routine. “He’d bike 20km, swim for an hour, then come to the office relaxed,” says Voll. Until recently, Fischer and Harald would go biking on 40-kilometre rides. “Sometimes he was behind in my slipstream, then we change positions,” says Fischer. “We were proud of the competitive speed and times we could achieve together.” He was a keen water polo player in his youth.
Fluent in English, French, and his native German, Harald read books in all three languages. He adored France – his wife is French – with its pastries and Brittany with its Hydrangeas. He said how the Hydrangeas in his garden struggled, yet on seeing the flourishing bands of colour in Brittany (pictured), he was determined to try again.
Legacy
When colleagues talk about Harald, they recount his warmth and character.
Uremovic recollects sharing an office and hearing Harald’s diplomatic charm on calls. These experiences have shaped his ability to connect better with others. He will also miss their talks: “To me, he was like a big brother.”
Harald made a deep impression during his short time at Ekinops. “We miss him,” says Quartier. “We will always have his memory at Ekinops.”
Fischer highlights their shared interests in life – sport and a love of France.”There was a lot of mutual understanding which we didn’t need to put into words,” says Fischer. “Sometimes, when you feel close to somebody, you don’t even need to talk.”
Fischer, who has watched Harald’s career over the years, laments a life cut short. “He was at the peak of his career,” he says. ”He brought to Ekinops all his experience of the last 20 to 30 years so this period was maybe the most impactful time of his life.”
Richter’s grief over his lost colleague is evident in his sombre tone and demeanour, perhaps the deepest tribute one can pay to someone held so dear.
OFC 2025: industry reflections
Gazettabyte is asking industry figures for their thoughts after attending the recent 50th-anniversary OFC show in San Francisco. Here are the first contributions from Huawei’s Maxim Kuschnerov, NLM Photonics’ Brad Booth, LightCounting’s Vladimir Kozlov, and Jürgen Hatheier, Chief Technology Officer, International, at Ciena.
Maxim Kuschnerov, Director of R&D, Huawei
The excitement of last year’s Nvidia’s Blackwell graphics processing unit (GPU) announcement has worn off, and there was a slight hangover at OFC from the market frenzy then.
The 224 gigabit-per-second (Gbps) opto-electronic signalling is reaching mainstream in the data centre. The last remaining question is how far VCSELs will go—30 m or perhaps even further. The clear focus of classical Ethernet data centre optics for scale-out architectures is on the step to 448Gbps-per-lane signalling, and it was great to see many feasibility demonstrations of optical signalling showing that PAM-4 and PAM-6 modulation schemes will be doable.
The show demonstrations either relied on thin-film lithium niobate (TFLN) or the more compact indium-phosphide-based electro-absorption modulated lasers (EMLs), with thin-film lithium niobate having the higher overall optical bandwidth.
Higher bandwidth pure silicon Mach-Zehnder modulators have also been shown to work at a 160-175 gigabaud symbol rate, sufficient to enable PAM-6 but not high enough for PAM-4 modulation, which the industry prefers for the optical domain.
Since silicon photonics has been the workhorse at 224 gigabits per lane for parallel single-mode transceivers, a move away to thin-film lithium niobate would affect the density of the optics and make co-packaged optics more challenging.
With PAM-6 being the preferred modulation option in the electrical channel for 448 gigabit, it begs the question of whether the industry should spend more effort on enabling PAM-6 optical to kill two birds with one stone: enabling native signalling in the optical and electrical domains would open the door to all linear drive architectures, and keep the compact pure-silicon platform in the technology mix for optical modulators. Just as people like to say, “Never bet against copper,” I’ll add, “Silicon photonics isn’t done until Chris Doerr says so.”
If there was one topic hotter than the classical Ethernet evolution, it was the scale-up domain for AI compute architectures. The industry has gone from scale-up in a server to a rack-level scale-up based on a copper backplane. But future growth will eventually require optics.
While the big data centre operators have yet to reach a conclusion about the specifications of density, reach, or power, it is clear that such optics must be disruptive to challenge the classical Ethernet layer, especially in terms of cost.
Silicon photonics appears to be the preferred platform for a potential scale-up, but some vendors are also considering VCSEL arrays. The challenge of merging optics onto the silicon interposer along with the xPU is a disadvantage for VCSELs in terms of tolerance to high-temperature environments.
Reliability is always discussed when discussing integrated optics, and several studies were presented showing that optical chips hardly ever fail. After years of discussing how unreliable lasers seem, it’s time to shift the blame to electronics.
But before the market can reasonably attack optical input-output for scale-up, it has to be seen what the adoption speed of co-packaged optics will be. Until then, linear pluggable optics (LPO) or linear retimed optics (LRO) pluggables will be fair game in scaling up AI ‘pods’ stuffed with GPUs.
Brad Booth, CEO of NLM Photonics
At OFC, the current excitement in the photonics industry was evident due to the growth in AI and quantum technologies. Many of the industry’s companies were represented at the trade show, and attendance was excellent.
Nvidia’s jump on the co-packaged optics bandwagon has tipped the scales in favour of the industry rethinking networking and optics.
What surprised me at OFC was the hype around thin-film lithium niobate. I’m always concerned when I don’t understand why the hype is so large, yet I have still to see the material being adopted in the datacom industry.
Vladimir Kozlov, CEO of LightCounting
This year’s OFC was a turning point for the industry, a mix of excitement and concern for the future. The timing of the tariffs announced during the show made the event even more memorable.
This period might prove to be a peak of the economic boom enabled by several decades of globalisation. It may also be the peak in the power of global companies like Google and Meta and their impact on our industry.
More turbulence should be expected, but new technologies will find their way to the market.
Progress is like a flood. It flows around and over barriers, no matter what they are. The last 25 years of our industry is a great case study.
We are now off for another wild ride, but I look forward to OFC 2050.
Jürgen Hatheier, Chief Technology Officer, International, at Ciena
This was my first trip to OFC, and I was blown away. What an incredible showcase of the industry’s most innovative technology
One takeaway is how AI is creating a transformative effect on our industry, much like the cloud did 10 years ago and smartphones did 20 years ago.
This is an unsurprising observation. However, many outside our industry do not realise the critical importance of optical technology and its role in the underlying communication network. While most of the buzz has been on new AI data centre builds and services, the underlying network has, until recently, been something of an afterthought.
All the advanced demonstrations and technical discussions at OFC emphasise that AI would not be possible without high-performance network infrastructure.
There is a massive opportunity for the optical industry, with innovation accelerating and networking capacity scaling up far beyond the confines of the data centre.
The nature of AI — its need for intensive training, real-time inferencing at the edge, and the constant movement of data across vast distances between data centres — means that networks are evolving at pace. We’re seeing a significant architectural shift toward more agile, scalable, and intelligent infrastructure with networks that can adapt dynamically to AI’s distributed, data-hungry nature.
The diversity of optical innovation presented at the conference ranged from futuristic Quantum technologies to technology on the cusp of mainstream adoption, such as 448-gigabit electrical lanes.
The increased activity and development around high-speed pluggables also show how critical coherent optics has become for the world’s most prominent computing players.
Books of 2024: Part 1
Gazettabyte asks industry figures to pick their notable reads during the year. Harald Bock, Jonathan Homa, and Maxim Kuschenrov kick off with their chosen books.
Harald Bock, Vice President Network Architecture, Infinera
I love reading but have not read as many books as I would have liked in recent years. I decided to change that in 2024.
My pick of fictional books this year was mainly classic science fiction after seeing the movie Dune Part 2 with my family. I read the book Dune by Frank Herbert, published in 1965, a while ago, and I wasn’t sure that the movies did the book justice.
My son advised me to launch myself into all five sequels of Dune, which kept me busy. While the sequels are for die-hard fans, I recommend the first of the books whether or not you’ve seen the movie. Frank Herbert’s modern and sophisticated thinking adds unconventional perspectives to up-to-date societal, environmental and political questions.
I went on to read Ray Bradbury’s The Martian Chronicles and H.G. Wells’ Time Machine, published in 1950 and 1895, respectively. The two books are fascinating as they are timeless and do not require any adaptation to modern times. They are classics of their genre.
I also found time for non-fictional books. I was looking for unconventional thoughts by unlikely authors to challenge my thinking.
One that adds to the discussions about sustainability is a book by Fred Vargas, a French author who normally writes crime fiction and is an archaeologist and historian. ‘L’humanité en péril: Virons de bord, toute !‘ was published as a follow-up to an older, shorter text by the same author read on the occasion of the conference on climate change COP24 in Paris in 2018. Surprisingly, the book does not yet exist in English.
Another interesting author is a professor of computer science, Katharina Zweig. Her books: Awkward Intelligence: Where AI Goes Wrong, Why It Matters, and What We Can Do about It and Die KI war’s: Von absurd bis tödlich: Die Tücken der künstlichen Intelligenz (‘It was the AI: From absurd to deadly: The pitfalls of artificial intelligence’, in German only to date) do a good job exploring considerations, boundary conditions, and limits of using AI systems in practical decision-making.
Jonathan Homa, Senior Director of Solutions Marketing at Ribbon Communications
I recommend a book I re-read this year: The Name of the Rose by Umberto Eco. As my wife points out, re-reading a book is its own recommendation.
This is an intricate and beautifully written murder mystery novel set in late medieval Europe. Through the eyes of the protagonist, Brother William of Baskerville, we begin to see glimpses of enlightenment. I also recommend the 1986 movie by the same name, starring Sean Connery.
Maxim Kuschnerov, director of R&D at Huawei
I had a light year of reading. One book I did read was Nuclear War: A Scenario by Annie Jacobsen which details the scenario of how a nuclear war would go down if someone started it. The answer: a surprisingly quick annihilation of humankind.
I also read Angela Merkel’s autobiography, Freedom: Memoirs 1954 – 2021 – that was published recently. I was hoping for more insight into her thinking when dealing with the immigration crisis or with Vladimir Putin, but the book added nothing that I didn’t already know about her. The book clarified how Angela Merkel was profoundly shaped in her upbringing by Eastern German communism and Russia.
The request to highlight my reads of 2024 made me think about what I have been reading this past year. Perhaps disappointingly, it turned out to be mostly not-noteworthy fiction.
ECOC 2024 industry reflections
Gazettabyte is asking industry figures for their thoughts after attending the recent 50th-anniversary ECOC show in Frankfurt. Here are the first contributions from Huawei's Maxim Kuschnerov, Coherent's Vipul Bhatt, and Broadcom's Rajiv Pancholy.
Maxim Kuschnerov, Director R&D, Optical & Quantum Communication Laboratory at Huawei.
At ECOC, my main interest concerned the evolution of data centre networking to 400 gigabits per lane for optics and electronics. Historically, the adoption of new optical line rates always preceded the serdes electrical interconnects but now copper cables are likely to drive much of the leading development work at 400 gigabit per lane.
Arista Networks argued that 448G-PAM6 works better for copper, while 448G-PAM4 is the better choice for optics – a recurring argument. While PAM6 signalling is certainly more suitable for longer copper cables, it will face even tougher challenges on the optical side with increasing reflection requirements in newly built, dusty data centres. Also, a linear drive option for future Ethernet will be imperative, given the DSP’s increasing share of the the consumption in pluggable modules. Here, a native 448G-PAM4 format for the serdes (the attachment unit interface or AUI) and optics looks more practical.
My most important takeaway regarding components was the initial feasibility of electro-absorption modulated lasers (EMLs) with a greater than 100GHz analogue bandwidth, presented by Lumentum and Mitsubishi publicly and other companies privately. Along with thin-film lithium niobate (TFLN) Mach–Zehnder modulators suited for Direct Reach (DR) applications with shared lasers, EMLs have historically offered low cost, small size and native laser integration.
For 1.6-terabit modules, everyone is waiting on the system availability of 224-gigabit serdes at a switch and network interface card (NIC) level. The power consumption of 1.6-terabit optical modules will improve with 3nm CMOS DSPs and native 200 gigabit per lane. Still, it gets into an unhealthy region where the network cable power consumption is in the same ballpark as the system function of switching. Here, the bet on LPO certainly didn’t pay off at 100 gigabits per lane and will not pay off at 200 gigabits per lane at scale. The question is whether linear receive optics (LRO)/ half-retimed approaches will enter the market. Technically, it’s feasible. So, it might take one big market player with enough vertical integration capability and a need to reduce power consumption to move the needle into this more proprietary, closed-system direction. Nvidia showcased their PAM4 DSP at the show. Just saying…
212G VCSELs are still uncertain. There is a tight initial deployment window to be hit if these high-speed VCSELS are to displace single-mode fibre-based optics at the major operators. Coherent’s results of 34GHz bandwidth are not sufficient and don’t look like something that could yet be produced at scale. Claims by some companies that a 400 gigabit per lane VCSEL is feasible sound hollow for now, with the industry crawling around the 30GHz bandwidth window.
Last but not least, co-packaged optics. For years, this technology couldn’t escape gimmick status. Certainly, reliability, serviceability, and testability of co-packaged optics using today’s methodology would make a deployment impractical. However, the big prize at 400 gigabit per lane is saving power – a significant operational expense for operators – something that is too attractive to ignore.
The targets of improving optics diagnostics, developing higher-performance dust-reflection DSP algorithms to deal with multi-path interference, adopting more resiliency to failure in the network, and introducing a higher degree of laser sparing are not insurmountable tasks if the industry sets its mind to them. Given the ludicrous goals of the AI industry, which is reactivating and rebranding nuclear power plants, a significant reduction in network power might finally serve a higher purpose than just building a plumber’s pipe.
Vipul Bhatt, Vice President of Marketing, Datacom Vertical, Coherent
ECOC 2024 was the most convincing testimony that the optical transceiver industry has risen to the challenge of AI’s explosive growth. There was hype, but I saw more solid work than hype. I saw demonstrations and presentations affirming that the 800-gigabit generation was maturing quickly, while preparations are underway for the next leap to 1.6 terabit and then 3.2 terabit.
This is no small feat, because the optics for AI is more demanding in three ways. I call them the three P’s of AI optics: performance, proliferation, and pace.
Performance because 200 gigabit PAM4 optical lanes must work with a low error rate at higher bandwidth. Proliferation because the drive to reduce power consumption has added new transceiver variants like linear packaged optics (LPO) and linear receive optics (LRO). And pace because the specifications of AI optics are evolving at a faster pace than traditional IEEE standards.
Rajiv Pancholy, Director of Hyperscale Strategy and Products, Optical Systems Division, Broadcom
As generative AI systems move to unsupervised, transformer-based parallel architectures, there is less time for resending packets due to data transmission errors. Improved bit error rates are thus required to reduce training times while higher interconnect bandwidth and data rates are needed to support larger GPU clusters. These compute networks are already moving to 224 gigabit PAM4 well before the previous generation at 112 gigabit PAM4 was allowed to reach hyperscale deployment volumes.
The problem is scalability with a high-radix supporting all-to-all connectivity. The power for a single rack of 72 GPUs is 120kW, and even with liquid cooling, this becomes challenging. Interconnecting larger scale-up and scale-out AI computing clusters requires more switching layers which increases latency.
Furthermore, after 224 gigabit PAM4, the losses through copper at 448 gigabit PAM4 make link distances from the ASIC too short. Moving to modulation schemes like PAM-6 or PAM-8 presents a problem for the optics, which would need to stay at 448 gigabit PAM4 to minimize crosstalk and insertion losses.
Supporting 448 gigabit PAM4 with optics then potentially requires new materials to be integrated into silicon, like thin-film lithium niobate (TFLN) and Barium Titanate (BaTiO3), electro-optic (EO) polymers, and III-V materials like Indium Phosphate (InP) and Gallium Arsenide (GaAs). So now we have a gearbox and, potentially, a higher forward error correction (FEC) coding gain is required, adding more power and latency before the signal even gets to the transmit-side optics.
There were 1.6-terabit OSFP transceivers operating with eight lanes of 212.5 gigabit PAM4 while vendors continue to work towards a 3.2-terabit OSFP-XD. With 32 x 3.2Tbps pluggables operating at 40W each, the optical interconnect power would be 1.3kW for a 102.4Tbps switch. And if you use 64 x 1.6Tbps OSFP at 25W each, the optical interconnect power will be eben higher, at 1.6kW. I wonder how linear pluggable optics can compensate for all the path impairments and reflections at high data rates from pluggable solutions. Perhaps you can relax link budgets, temperature requirements, and interoperability compliance.
The best session this year was the last ECOC Market Focus panel on the Tuesday, which kept everyone a bit longer before they could figure out where in Frankfurt Oktoberfest beer was on tap. The panel addressed “Next-Gen Networking Optics like 1.6T or 3.2T”. All but one of the participants discussed the need and a migration to co-packaged optics, which we at Broadcom first demonstrated in March 2022.
It was great to also present at the ECOC Market Focus forum. My presentation was titled “Will you need CPO in 3 years?” Last year in Glasgow, I gave a similar presentation: “Will you need CPO in 5 years?”
OFC 2024 industry reflections
Gazettabyte is asking industry figures for their thoughts after attending the recent OFC show in San Diego. In particular, what developments and trends they noted, what they learned and what, if anything, surprised them. Here are the first responses from Huawei, Drut Technologies and Aloe Semiconductor.
Maxim Kuschnerov, Director R&D, Optical & Quantum Communication Laboratory at Huawei.
Some ten years ago datacom took the helm of the optical transceiver market from legacy telecom operators to command a much larger volume of short-reach optics and extend its vision into dense wavelength division multiplexing (DWDM).
At OFC, the industry witnessed another passing-of-the-torch moment as Nvidia took over the dominant position in the optics market where AI compute is driving optical communication. The old guard of Google is now following while others are closely watching.
Nvidia’s Blackwell NVL72 architecture was the talk of the conference and its exclusive reliance on passive copper cables for intra-rack GPU-to-GPU interconnects dampened the hopes of Wall Street optics investors at the show.
Since the copper backplane is using 224-gigabit serdes, last year’s hot topics of 100 gigabit-based linear pluggable optics or dense optical interconnects based on 16×100 gigabits suddenly felt dated and disconnected from where the market already is. It is also in no shape to respond to where the compute market is rapidly going next: 400-gigabit-per-lane signalling.
Here, the main question is which type of connectivity for the GPU scale-up in the intra-rack domain would be employed and whether this might be the crossover point to go to optical cables? But as often is the case in the optical business, one should never fully bet against CMOS and copper.
The long-term evolution of AI compute will impact optical access and this was also a theme of some of the OFC panels.
6G is envisioned to be the first wireless network supporting devices primarily, not humans, and it’s fair to assume that a majority of those distributed devices will be AI-enabled. Since it will be uneconomical to send the raw training or inference bandwidth to the network core, the long term evolution of AI compute might see a regionalisation and a distribution towards the network edge, where there would be a strong interdependence of 6G, fronthaul/ backhaul & metro edge networks, and the AI edge compute cloud.
While a majority of coherent PON presentations failed to quantify the market driver for introducing the more expensive technology in future access networks, AI-data powered 6G fronthauling over installed optical distribution networks will drive the bandwidth need for this technology, while residential fibre-to-the-home – “PON for humans” – can still evolve to 200 gigabit using low cost intensity modulation direct detection (IMDD) optics.
The times are over where the talk of cheaper datacom ZR optics dominated the show and commanded attendance at the Market Watch sessions. Don’t misunderstand, the step to 1600ZR is technologically important and market-relevant, but since coherent doesn’t have “AI” written all over it, the ZR evolution was more a footnote of the exhibition. However, in a necessary move away from electro-absorption-modulated lasers (EMLs), 400-gigabit-per-lane optics for intensity modulation direct detection will share similar Mach-Zehnder modulator structures as coherent optics.
Thus, start-ups crowding the thin-film lithium niobate modulator market in the US, Europe and China are going for both: the coherent and the intensity modulation direct detection dollar.
Meanwhile, the established silicon photonics ecosystem will have to wrap its head around what their value-add in this domain will be since silicon photonics would be just the carrier of other materials enabling lasers, modulators and photodetectors.
Bill Goss, CEO of Drut Technologies
The last time I attended OFC, the conference was in Los Angeles at the Staples Center.
One thing I found super interesting at this year’s event was the number of companies working on optically-connected memory solutions. But the biggest noteworthy item to us was a number of presentations on using optical circuit switching (OCS) for AI/ML workloads.
Nvidia and some universities presented projects using OCS in the data centre and Coherent actually showed a new 300×300 switch in their booth. There also seemed to be a feeling that the world has been waiting on co-packaged optics for years.
One thing evident in talking with optical companies that typically focus on service provider networks, is that they all want to get inside the data centre. That is where the big market explosion is going to be in the next decade and companies are thinking about how to gain share in the data centre with optical solutions.
You could almost feel the gloom around service provider capital expenditure and the companies that normally play in this market are looking at all the spending going on inside the data centre and trying to figure out how to access this spend.
Drut Technologies did not exhibit at OFC. Instead, we used the show to listen to presentations and talk to suppliers and customers. Surprises were the amount of pluggable optics available.
Walking through the show floor, it seemed like a sea of pluggables and I had multiple meetings with companies looking to put coherent optics inside the data centre. Visually too, the amount of pluggables was noticeable.
I was also surprised at the absence of certain companies. It seems companies opted for a private meeting room rather than a booth. I do not know what that means, if anything, but if the trend continues, the show floor is going to be half-filled with private meeting spaces. It will be like walking through a maze of white walls.
I was not surprised with all the AI excitement, but the show did not seem to have a lot of energy.
Chris Doerr, CEO of Aloe Semiconductor
The first most noteworthy trend of this OFC was the acceleration of pluggable module data rates. There were demonstrations of 1.6-terabit pluggables by almost every module vendor. This was supposed to be the year of 800 gigabit not 1.6 terabit.
Digging into it more, most of the demonstrated 1.6 terabit modules were not fully operational – the receiver was not there, all the channels not running simultaneously, etc. – but some EML-based modules were complete.
The second most noteworthy trend was supply constraint and the subsequent driving of new technology. For example, it was said that Nvidia bought up all the VCSEL supply capacity. This is driving up VCSEL prices and seems to be allowing a surge of silicon photonics in the lower speed markets that were previously thought to be done and closed, such as active optical cables. There was an increasing polarity in opinion on linear pluggable optics, with opposing opinions by well-known technologists.
It seems that Nvidia is already deploying 100 gigabit per lane linear pluggable optics, and Arista will be deploying it soon. For 200 gigabit per lane, it seems the trend is to favour half-linear pluggable optics, or linear receive optics (LRO), in which the transmit is still retimed.
Large-scale co-packaged optics (not to be confused with small-scale CPO of a coherent ASIC and coherent optics) was exhibited by more vendors this year. It seems very little, if any, is deployed. Large-scale CPO is inevitable, but it on a significantly slower time scale than previously thought.
For 200 gigabit per lane, there were many demonstrations using EMLs and quite a few using silicon photonics. Most of the silicon photonics demonstrations seemed to require driver ICs to overcome the reduced modulation efficiency, sacrificed to achieve the higher bandwidth. Consequently, most companies appear to be throwing in the towel on silicon photonics for 200 gigabaud (GBd) applications, instead moving toward indium phosphide and thin-film LiNbO3 (TFLN). This is surprising.
This author strongly believes in the trend usually followed by silicon electronics in that innovation will allow silicon photonics to achieve 200GBd. It is unreasonable to expect indium phosphide or TFLN to meet the volumes, density, and pricepoints required for 3.2-terabit modules and beyond.
There is no widely accepted solution for 400-gigabit-per-lane intensity modulation direct detection. Proposals include two wavelengths x 200 gigabit, going for 200GBd early, and dual-polarization intensity modulation direct detection.
There was significant discussion about optoelectronic interposers, with start-ups LightMatter and Celestial AI receiving large funding in this area. However, the end customers do not seem to have a need for this technology, so it is unclear where it is headed.
OFC was highly noteworthy this year, driven by the surging demand for high-performance computing interconnects. Probably the biggest takeaway is the amount of uncertainty and polarised views, including linear pluggable optics, silicon-photonic’s future, and optoelectronic interposers.
ECOC 2023 industry reflections
Gazettabyte is asking industry figures for their thoughts after attending the recent ECOC show in Glasgow. In particular, what developments and trends they noted, what they learned and what, if anything, surprised them. Here are the first responses from BT, Huawei, and Teramount.
Andrew Lord, Senior Manager, Optical Networks and Quantum Research at BT
I was hugely privileged to be the Technical Co-Chair of ECOC in Glasgow, Scotland and have been working on the event for over a year. The overriding impression was that the industry is fully functioning again, post-covid, with a bumper crop of submitted papers and a full exhibition. Chairing the conference left little time to indulge in content. I will need to do my regular ECOC using the playback option. But specific themes struck me as interesting.
There were solid sessions and papers around free space optics, including satellite. The activities here are more intense than we would typically see at ECOC. This reflects a growing interest and the specific expertise within the Scottish research community. Similarly, more quantum-related papers demonstrated how quantum is integrating into the mainstream optical industry.
I was impressed by the progress towards 800-gigabit ZR (800ZR) pluggables in the exhibition. This will make for some interesting future design decisions, mainly if these can be used instead of the increasingly ubiquitous 400 gigabit ZR. I am still unclear whether 800-gigabit coherent can hit the required power consumption points for plugging directly into routers. The costs for these plugs, driven by volumes, will have a significant impact.
I also enjoyed a lively and packed rump session debating the invasion of artificial intelligence (AI) into our industry. I believe considerable care is needed, particularly where AI might have a role in network management and optimisation.
Maxim Kuschnerov, Director R&D at Huawei
ECOC usually has fewer major announcements than the OFC show. But ECOC was full of technical progress this time, making the OFC held in March seem a distant memory.
What was already apparent in September at the CIOE in Shenzhen was on full display on the exhibition floor in Glasgow: the linear drive pluggable optics (LPO) trend has swept everyone off their feet. The performance of 100-gigabit native signalling using LPO can not be ignored for single-mode fibre and VCSELs.
Arista gave a technical deep-dive at the Market Focus with a surprising level of detail that went beyond the usual marketing. There was also a complete switch set-up at the Eoptolink booth, and the OIF interop demonstration.
While we must wait for a significant end user to adopt LPO, it begs the question: is this a one-off technological accident or should the industry embrace this trend and have research set its eyes on 200 gigabits per lane? The latter would require a rearchitecting of today’s switches, a more powerful digital signal processor (DSP) and likely a new forward error corrections (FEC) scheme, making the weak legacy KP4 for the 224-gigabit serdes in the IEEE 802.3dj look like a poor choice.
There was less emphasis on Ethernet 1.6 terabits per second (Tb/s) interfaces with 8x200G optical lanes. However, the arrival of a second DSP source with better performance was noted at the show.
The module power of 1.6-terabit DR8 modules showed no significant technological improvement compared with 800Gbps DSP-based modules and looked even more out of place when benchmarking against 800G LPO pluggables. Arista drove home that we can’t continue increasing the power consumption of the modules at the faceplate despite the 50W QSFP-DD1600 announcement.
The same is true for coherent optics.
Although the demonstration of the first 800ZR live modules was technically impressive, the efficiency of the power per bit hardly improved compared to 400ZR, making the 1600ZR project of OIF look like a tremendous technological challenge.
To explain, a symbol rate of 240 gigabaud (GBd) will drive the optics for 1600ZR. Using 240Gbaud with two levels per symbol to create 16QAM over two dimensions is a 400Gbps net rate or 480Gbps gross rate electrical per lane, albeit very short reach. Coherent has four lanes – 2 polarisations & in-phase and quadrature – to deliver four by 400G or 1.6Tbps. This is like what we have now: 200G on the optical side of 1.6T 8x200G PAM4 and 4x200G on 800ZR, while the electrical (longer reach) host still uses 100 gigabits per lane.
The industry will have to analyse which data centre scenarios direct detection will be able to cover with the same analogue-to-digital & digital-to-analogue converters and how deeply coherent could be driven within the data centre.
ECOC also featured optical access evolution. With the 50G FTTx standard completed with components sampling at the show and products shipping next year, the industry has set its eyes on the next generation of very high-speed PON.
There is some initial agreement on the technological choice for 200 gigabits with a dual-lambda non-return to zero (NRZ) signalling. Much of the industry debate was around the use cases. It is unrealistic to assume that private consumers will continue driving bandwidth demand. Therefore, a stronger focus on 6G wireless fronthaul or enterprise seems a likely scenario for point-to-multi-point technology.
Hesham Taha, CEO of Teramount
Co-packaged optics had renewed vigour in ECOC, thanks partly to the recent announcements of leading foundries and other semiconductor vendors collaborating in silicon photonics.
One crucial issue, though, is that scalable fibre assembly remains an unsolved problem that is getting worse due to the challenging requirements of high-performance systems for AI and high-performance computing. These requirements include a denser “shoreline” with a higher fibre count and a denser fibre pitch, and support for an interposer architecture with different photonic integrated component (PIC) geometries.
Despite customers having different requirements for co-packaged optics fibre assembly, detachable fibres now have wide backing. Having fibre ribbons that can be separated from the co-packaged optics packaging process increases manufacturing yield and reliability. It also allows the costly co-packaged optics-based servers/ switches to be serviced in the field ro replace faulty fibre.
Our company, Teramount, had an ECOC demo showing the availability of such a detachable fibre connector for CPO, dubbed Teraverse.
It is increasingly apparent that the solution for a commercially viable fibre assembly on chip lies with a robust manufacturing ecosystem rather than something tackled by any one system vendor. This fabless model has proven itself in semiconductors and must be extended to silicon photonics. This will allow each part of the production chain – IC designers, foundries, and outsourced semiconductor assembly and test (OSAT) players – to focus on what they do best.
Data centre photonics - an ECOC report
- ECOC 2022 included talks on optical switching and co-packaged optics.
- Speakers discussed optical switching trends and Google’s revelation that it has been using optical circuit switching in its data centres.
- Nvidia discussed its latest chips, how they are used to build high-performance computing systems, and why optical input-output will play a critical role.
Co-packaged optics and optical switching within the data centre were prominent topics at the recent ECOC 2022 conference and exhibition in Basel, Switzerland.
There were also two notable data centre announcements before ECOC.
Tencent announced it would adopt Broadcom’s Humboldt design, a hybrid co-packaged optics version of the Tomahawk 4 switch chip, in its data centres. Tencent is the first hyperscaler to announce it is adopting co-packaged optics.
Google also revealed its adoption of optical circuit switching in its data centres. Google made the revelation in a paper presented at the Sigcomm 2022 conference held in Amsterdam in August.
Optical circuit switching
Google rarely details its data centre architecture, but when it does, it is usually at Sigcomm.
Google first discussed a decade of evolution of its ‘Jupiter’ data centre architecture in a paper at Sigcomm in 2015.
This year, Google gave an update revealing that it has been using optical circuit switching in its data centres for the past five years. As a result, Google can scale its data centre more efficiently using a reconfigurable optical layer.
Upgrading a data centre’s network is much more complex than upgrading servers and storage. Moreover, a data centre is operational far longer than each generation of equipment. It is thus hard for a data centre operator to foresee how equipment and workloads will evolve over the data centre’s lifetime, says Google.
Google would pre-deploy the spine layer when it started operating a data centre. For Google’s Jupiter architecture, 64 spine blocks, each using 40 gigabit-per-second (Gbps) links, would be deployed. Then, Google added newer aggregation blocks with 100Gbps links. But the hyperscaler could not fully benefit due to the pre-existing 40Gbps spine links.
Google wanted to avoid touching the spine switches. A partial upgrade would have limited benefits, while fully upgrading the spine would take months and be hugely disruptive and costly.
Google’s first solution introduced a MEMS-based optical circuit switching layer between the aggregation and spine blocks.
The MEMS-based switch is data rate agnostic and can support multiple generations of optical modules. The switch’s introduction also allowed Google to add new spine blocks alongside new aggregation blocks; the hyperscaler no longer had to pre-deploy the spine.
At some point, Google decided that for new data centre builds, it would use optical circuit switching only and remove the spine layer of electrical switches.
Adopting optical circuit switch-based interconnect changes Google’s data centres from a clos to a direct-connect architecture. However, not all paths are direct; some take two hops to link aggregation blocks.
Google has developed sophisticated control software to best exploit the direct connectivity for traffic flows.
The software also adapts the network topology – the optical links between the aggregation blocks and their capacities. Such topology changes occur every few weeks, with the system first learning the nature of the traffic and workloads.
Removing the spine layer and replacing it with optical circuit switches has reduced Google’s data centre networking costs by 30 per cent and power consumption by 41 per cent.
ECOC reflections about Google’s optical switch adoption
There was much discussion at ECOC of Google’s use of optical circuit switching in its data centres.
S.J. Ben Yoo, a distinguished professor at the University of California, Davis, gave an ECOC talk about new trends in optical switching. “These are expected future trends,” he said. “I don’t have a crystal ball.”
Prof. Ben Yoo stressed the difficulty of scaling up and scaling out data centre networking architectures in the era of artificial intelligence workloads.
He described co-packaged optics as ‘Trend 0’ because it only delivers bandwidth (input-output capacity).
In contrast, introducing a reconfigurable optical switching layer on top of electrical aggregation switches is the first trend in optical switching. This is what Google has done with its optical circuit switch.
The next development in the data centre, says Ben Yoo, will be the introduction of photonic integrated circuit-based optical switching.
Huawei’s Maxim Kuschnerov, in his ECOC talk, said optical switching in the data centre would only grow in importance.
“Are there use cases where we can use optical switching and what are they?” he said. “I like to take a use-case perspective and find a technology that fulfils that use case.”
His view is that with the classical clos architecture, you can’t just rip out a single layer of electrical switches and replace it with optical ones. “There is a reason why you need all these switches and aggregation functionality,” says Kuschnerov.
Kuschnerov views Google’s optical circuit switching as nothing more than an automated patch panel.
“This is not the optical switch which is the saviour of future data centres,” he says.
Mark Filer, optical network architect, systems and services infrastructure at Google, in an ECOC tutorial detailing how Google uses and benefits from standards, multi-source agreements and open-source developments, was asked about Google’s custom optical switch.
How could Google explain such a custom design if the hyperscaler is such a proponent of open standards? And would Google consider contributing its optical circuit switch and software design to an open community framework?
“My guess is over time, we may see that it finds its way into some kind of open framework,” said Filer, adding that right now, Google sees its optical circuit switch as delivering a competitive advantage.
Co-packaged optics
Benjamin Lee, a senior research scientist at Nvidia, in his ECOC address, discussed the high-performance computing market and the role graphics processing units (GPUs) play in accelerating artificial intelligence and machine learning tasks.
Nvidia not only develops processors, GPUs and data processing unit ICs but also networking silicon and systems that the company uses to make high-performance computing systems.
Lee’s talk addressed the role optical interconnect will play in ensuring continuing scaling of high-performance GPU-based computing systems.
Scaled systems
Nvidia’s latest GPU, announced earlier this year, is the 80-billion-transistor Hopper H100. The H100 deliver a six-fold improvement in throughput compared to Nvidia’s existing A100 GPU announced in 2020.
The Hopper is Nvidia’s first GPU that uses the latest generation of stacked DRAM memory, known as high bandwidth memory 3 (HBM3). In addition, Hopper also uses Nvidia’s fourth-generation NVlink interface.
Eight H100 GPUs fit within Nvidia’s DGX box, as do four Nvidia NVSwitches used to interconnect the GPUs. In addition, an Nvidia Superpod connects 32 DGX nodes – 256 GPUs – using an external tier of NVSwitches.
“A paradigm shift we’re seeing is that switched interconnect is becoming important for scale-up,” said Lee. “So when we want to make the node more computationally powerful, those switches are being put inside the box to connect the GPUs.”
Switch ASIC bandwidths are consistently improving, with 51.2-terabit switch silicon being state-of-the-art. But despite such progress, the scaling is insufficient to keep up with bandwidth requirements, said Lee.
Switch ASIC power consumption is also rising, with advanced CMOS scaling having less impact on designs. Lee foresees switch ASICs consuming 2kW if current trends continue.
In turn, ASIC input-output (I/O) accounts for an increasing portion of the chip’s overall power consumption.
This is true for Nvidia’s GPUs and switch chips, so any I/O technology developed for switching will also benefit its GPUs.
Thus, Nvidia sees optical I/O as the key to scaling the processing performance of its ASICs and computing systems.
I/O metrics
Lee outlined various metrics when discussing optical I/O:
- the electrical interfaces used between the ASIC and optics, and their reach
- the power consumption of the module (the chip, and the chip and optics)
- the system power (of the line card or platform)
- interface density: the capacity exiting a millimetre of surface in terabits-per-second-per-mm (Tbps/mm)
For a system using a 102.4-terabit switch IC, half the power is consumed by the ASIC and half by the edge-board pluggable optics. Here the OIF’s long reach (LR) interface links the two.
The chip’s electrical interfaces consume 4.5 to 6.5 picojoule-per-bit (pJ/b) such that the total switch IC I/O power consumed is 450W.
The next step is co-packaged optics. Here, optical chiplets are placed closer to the ASIC (100mm away) such that the OIF’s lower power XSR (extra short reach) interface can be used that consumes 1.24-1.7pJ/s, says Lee.
Again taking a module view, Nvidia views the co-packaged design as comprising two electrical interfaces (the XSR interface between the chip and optical chiplets either side) and one optical interface.
This equates to 250W per chip module, a modest power saving at the chip module level but a significant power saving at the system level, given the optics is now part of the module.
However, bandwidth density is 475-870Gbps/mm, and for beyond 100-terabit switches, a further fourfold improvement is needed: 2Tbps/mm and, ultimately, 10Tbps/mm.
Just achieving a 2Tb/s/mm interface density will be challenging, says Lee.
For that, 2.5D co-packaged optics will be needed with the ASIC and chiplets sharing a silicon interposer that enables higher wire densities.
2.5D integration is already an established technology in the semiconductor industry; Nvidia has been using the technology for its GPUs since 2016.
The technology enables much closer coupling between the ASIC and optics (some 1mm), resulting in sub 1pJ/bit. Nvidia cites research showing a 0.3pJ/b has already been achieved.
Scaling I/O
Lee outlined all the ways I/O can be scaled.
Baud rate is one approach, but the energy efficiency diminishes as the symbol rate increases, from 50 to 100 to 200 gigabaud.
Modulation is another approach, moving from non-return-to-zero to 4-level pulse amplitude modulation (PAM-4) and even higher PAM schemes. The challenge is that the signal-to-noise ratio diminishes the higher the PAM scheme, requiring additional digital signal processing which, in turn, consumes more power.
Another technique, polarization, can be used to double the data rate. Then there is the spatial domain. Here, tighter pitches can be used, says Lee, moving from 250, 127 and even 80 microns before other approaches are needed. These include multi-core fibre, waveguide fan-outs and even bidirectional optics (what Google uses for its optical circuit switch ports, to save on fibre and port count).
All these spatial approaches require considerable development and operational costs, says Lee.
The most promising way to boost throughput and increase interface density is using wavelength division multiplexing (WDM).
Nvidia has produced several generations of test chips that use wavelength parallelism in the O-band based on micro-ring resonators.
Nvidia’s steer
Micro-ring resonator technology already supports 100Gbps modulation rates. The optical circuit is also compact, energy-efficient and supports wavelength scaling.
Lee also outlined other key technologies that will be needed, each bringing their own challenges. One is the external laser source, another is advanced packaging.
Nvidia believes that for future generations of ASICs, dense WDM mirror-ring modulated links offer the most promising approach to meeting both low power and the massive interface density improvements that will be needed.
This will require low-cost lasers while packaging remains a severe challenge.
2.5D integration is going to be an important step in the evolution of switch interconnect, concluded Lee.
ECOC 2022 Reflections - Part 1
Gazettabyte is asking industry and academic figures for their thoughts after attending ECOC 2022, held in Basel, Switzerland. In particular, what developments and trends they noted, what they learned, and what, if anything, surprised them.
In Part 1, Infinera’s David Welch, Cignal AI’s Scott Wilkinson, University of Cambridge’s Professor Seb Savory, and Huawei’s Maxim Kuschnerov share their thoughts.
David Welch, Chief Innovation Officer and Founder of Infinera
First, we had great meetings. It was exciting to be back to a live, face-to-face industry event. It was also great to see strong attendance from so many European carriers.
Point-to-multipoint developments were a hot topic in our engagements with service providers and component suppliers. It was also evident in the attendance and excitement at the Open XR Forum Symposium, as well as the vendor demos.
We’re seeing that QSFP-DD ZR+ is a book-ended solution for carriers; interoperability requirements are centred on the CFEC (concatenated or cascaded FEC) market; oFEC (Open FEC) is not being deployed.
Management of pluggables in the optical layer is critical to their network deployment, while network efficiency and power reduction are top of mind.
The definition of ZR and ZR+ needs to be subdivided further into ZR – CFEC, ZR+ – oFEC, and ZR+-HP (high performance), which is a book-ended solution.
Scott T. Wilkinson, Lead Analyst, Optical Components, Cignal AI.
The show was invaluable, given this was our first ECOC since Cignal AI launched its optical components coverage.
Coherent optics announcements from the show did not follow the usual bigger-faster-stronger pattern, as the success of 400ZR has convinced operators and vendors to look at coherent at the edge and inside the data centre.
100ZR for access, the upcoming 800ZR specifications from the OIF, and coherent LR (coherent designed for 2km-10km) will revolutionise how coherent optics are used in networks.
Alongside the coherent announcements were developments from the direct-detect vendors demonstrating or previewing key technologies for 800 Gigabit Ethernet (GbE) and 1.6 Terabit Ethernet (TbE) modules.
800GbE is nearly ready for prime time, awaiting completion of systems based on the newest 112 gigabit-per-second (Gbps) serialiser-deserialiser (serdes) switches. The technology for 224Gbps serdes is just starting to emerge and looks promising for products in late 2024 or 2025.
While there were no unexpected developments at the show, it was great to compare developments across the industry and understand the impact of supply chain issues, operator deployment plans, and any hints of oversupply.
We came away optimistic about continued growth in optical components shipments and revenue into 2023.
Seb Savory, Professor of Optical Fibre Communication, University of Cambridge
My overwhelming sense from ECOC was it was great to be meeting in person again. I must confess I was looking at logistics as much as content with a view to ECOC 2023 in Glasgow where I will be a technical programme committee chair.
Maxim Kuschnerov, Director of the Optical and Quantum Communications Laboratory at Huawei
In the last 12 months, the industry has got more technical clarification regarding next-generation 800ZR and 800LR coherent pluggables.
While 800ZR’s use case seems to be definitely in the ZR+ regime, including 400 gigabit covering metro and long-haul, the case for 800LR is less clear.
Some proponents argue that this is a building block toward 1.6TbE and the path of coherence inside the data centre.
Although intensity-modulation direct detection (IMDD) faces technical barriers to scaling wavelength division multiplexing to 8×200 gigabit, the technological options for beyond 800-gigabit coherent aren’t converging either.
In the mix are 4×400 gigabit, 2×800 gigabit and 1×1.6 terabit, making the question of how low-cost and low-power coherent can scale into data centre applications one of the most interesting technical challenges for the coming years.
Arista continues making a case for a pluggable roadmap through the decade based on 200-gigabit serdes.
With module power envelopes of around 40W at the faceplate, it shows the challenge that the industry is facing and the case co-packaged optics is trying to make.
However, putting all the power into, or next to, the switching chip doesn’t make the cooling problem any less problematic. Here, I wonder if Avicena’s microLED technology could benefit next-generation chip-to-chip or die-to-die interconnects by dropping the high-speed serdes altogether and thus avoiding the huge overhead current input-output (I/O) is placing on data centre networking.
It was great to see the demo of the 200-gigabit PAM-4 externally modulated laser (EML) at Coherent’s booth delivering high-quality eye diagrams. The technology is getting more mature, and next year will receive much exposure in the broader ecosystem.
As for every conference, we have seen the usual presentations on Infinera’s XR Optics. Point-to-multipoint coherent is a great technology looking for a use case, but it is several years too early.
At ECOC’s Market Focus, Dave Welch put up a slide on the XR ecosystem, showing several end users, several system OEMs and a single component vendor – Infinera. I think one can leave it at this for now without further comment.
Huawei sets transmission record with new modulator
Coherent discourse: Part 1
A paper from Huawei and Sun Yat-Sen University in the January issue of the Optica journal describes a thin-film lithium niobate modulator. The modulator enabled a world-record coherent optical transmission, sending nearly 2 terabits of data over a single wavelength.
Much of the industry’s focus in recent years has been to fit coherent optical technology within a pluggable module.
Such pluggables allow 400-gigabit coherent interfaces to be added to IP routers and switches, serving the needs of the data centre operators and telecom operators.
But research labs of the leading optical transport vendors continue to advance high-end coherent systems beyond 800-gigabit-per-wavelength transmissions.
Optical transport systems from Ciena, Infinera and Huawei can send 800-gigabit wavelengths using a symbol rate of 96-100 gigabaud (GBd).
Acacia Communications, part of Cisco, detailed late last year the first 1.2-terabit single-wavelength coherent pluggable transceiver that will operate at 140GBd, twice the symbol rate of 400-gigabit modules such as 400ZR.
Now Huawei has demonstrated in the lab a thin-film lithium niobate modulator that supports a symbol rate of 220GBd and beyond.
Maxim Kuschnerov, director of the optical and quantum communications laboratory at Huawei, says the modulator has a 110GHz 3dB bandwidth but that it can be operated at higher frequencies, suggesting a symbol rate as high as 240GBd.
Thin-film lithium niobate modulator
Huawei says research is taking place into new materials besides the established materials of indium phosphide and silicon photonics. “It is a very exciting topic lately,” says Kuschnerov.
He views the demonstrated thin-film lithium niobate optical modulator as disruptive: “It can cover up several deficiencies of today’s modulators.”
Besides the substantial increase in bandwidth – the objective of any new coherent technology – the modulator has performance metrics that benefit the coherent system such as a low driving voltage and low insertion loss.
A driving voltage of a modulator is a key performance parameter. For the modulator, it is sub-1V.
The signal driving the modulator comes from a digital-to-analogue (D/A) converter, part of the coherent digital signal processor (DSP). The D/A output is fed into a modulator driver. “That [driver] requires power, footprint, and increases the complexity of integrating the [modem’s] modules tighter,” says Kuschnerov.
The modulator’s sub-1V drive voltage is sufficiently small that the DSP’s CMOS-based D/A can drive it directly, removing the modulator driver circuit that also has bandwidth performance limitations. The modulator thus reduces the transmitter’s overall cost.
The low-loss modulator also improves the overall optical link budget. And for certain applications, it could even make the difference as to whether optical amplification is needed.
“The modulator checks the box of very high bandwidth,” says Kuschnerov. “And it helps by not having to add a semiconductor optical amplifier for some applications, nor needing a driver amplifier.”
One issue with the thin-film modulator is its relative size. While not large – it has a length of 23.5mm – it is larger than indium phosphide and silicon photonics modulators.
1.96-terabit wavelength
Huawei’s lab set-up used a transmit coherent DSP with D/As operating at 130 Giga-samples-per-second (GS/s) to drive the modulator. The modulation used was a 400-quadrature amplitude modulation (400-QAM) constellation coupled with probabilistic constellation shaping.
A 10 per cent forward error correction scheme was used such that, overall, 1.96-terabits per second of data was sent using a single wavelength.
The D/A converter was implemented in silicon germanium using high-end lab equipment to generate the signal at 130GS/s.
“This experiment shows how much we still need to go,” says Kuschnerov. “What we have done at 130GBd shows there is a clear limitation with the D/A [compared to the 220GBd modulator].”
Baud-rate benefits
Increasing the baud rate of systems is not the only approach but is the favoured implementation choice.
What customers want is more capacity and reducing the cost per bit for the same power consumption. Increasing the baud rate decreases the cost and power consumption of the optical transceiver.
By doubling the baud rate, an optical transceiver delivers twice the capacity for a given modulation scheme. The cost per bit of the transceiver decreases as does the power consumed per bit. Instead of two transceivers and two sets of components, one transceiver and one set are used instead.
But doubling the baud rate doesn’t improve the optical system’s spectral efficiency since doubling the baud rate doubles the channel width. That said, algorithmic enhancements are added to each new generation of coherent modem but technically, the spectral efficiency practically no longer improves.
Huawei acknowledges that while the modulator promises many benefits, all the coherent modem’s components – the coherent ASIC, the D/A and analogue-to-digital (D/A) converters, the optics, and the analogue circuitry – must equally scale. This represents a significant challenge.
Kuschnerov says optical research is finding disruptive answers but scaling performance, especially on the electrical side, remains a critical issue. “How do you increase the D/A sampling rates to match these kinds of modulator technologies?” he says. “It is not straightforward.”
The same is true for the other electrical components: the driver technologies and the trans-impedance amplifier circuits at the receiver.
Another issue is combining the electrical and optical components into a working system. Doubling the signalling of today’s optical systems is a huge radio frequency design and packaging challenge.
But the industry consensus is that with newer CMOS processes and development in components and materials, doubling the symbol rate again to 240GB will be possible.
But companies don’t know – at least they are not saying – what the upper symbol rate limit will be. The consensus is that increasing the baud rate will end. Then, other approaches will be pursued.
Kuschnerov notes that if a 1.6-terabit transceiver could be implemented using a single wavelength or with eight 200Gbps ones with the same spectral performance, cost, footprint and power consumption, end users wouldn’t care which of the two were used.
However, does optics enable such greater parallelism?
Kuschnerov says that while decades of investment has gone into silicon photonics, it is still not there yet.
“It doesn’t have the cost-effectiveness at 16, 32 or 64 lanes because the yield goes down significantly,” he says. “We as an industry can’t do it yet.”
He is confident that, soon enough, the industry will figure out how to scale the optics: “With each generation, we are getting better at it.”
Coherent engineers will then have more design options to meet the system objectives.
And just like with microprocessors, it will no longer be upping the clock frequency but rather adopting parallel processing i.e. multiple cores. Except, in this case, it will be parallel coherent optics.
Books read in 2021: Part 1
Each year Gazettabyte asks industry figures to pick their reads of the year. Paul Brooks and Maxim Kuschnerov kick off this year’s recommended reads.
Dr. Paul Brooks, Optical Transport Director, VIAVI Solutions
Having spent a very happy time serving in the Royal Navy, I am always reading about all things connected with its history.
As a young midshipman, I managed to sleep through many of the history lessons at BRNC Dartmouth so I am using my spare time to catch up on the lessons I missed all those years ago.
One book which I have very much enjoyed this year has been Stephen Taylor’s Sons of the Waves: The Common Seaman in the Heroic Age of Sail.
While many books are written about major figures such as Nelson and Blake, the ordinary sailor with his robustness, loyalty and sense of duty was the key element in the success of the Royal Navy.
This well-researched book is a joy to read as it brings to life the heroic men. I must confess I did hum ‘Heart of Oak’ as I reached for my tot of rum as I read about the jolly Jack Tar on the Victory at Trafalgar!
For any student of history, and indeed anyone interested in social history, this is one for your Christmas list.
Dr. Maxim Kuschnerov, Director of the Optical & Quantum Communications Laboratory
No Rules Rules: Netflix and the Culture of Reinvention, by Reed Hastings and Erin Meyer, got good press last year, so when I saw it at the airport, it was a no brainer to get it.
The book offers a radical approach to management, focusing on totally open feedback and the removal of most controls, whether it’s the lack of vacation policy (take as much as you want) or the absence of higher approvals for most business dealings. Salary adjustments are governed by external market references and not internal processes, which is generally not a bad thing.
Naturally, looking at this corporate culture through the glasses of a German dependency of a Chinese company makes for a big contrast and it would be hard to imagine a German company functioning without any kind of rules. But what the book achieves is to shift the normal operational bias towards a more modern view of team management and it helped me to make adjustments in everyday work, changing the way that I interpreted my role within my team.
This brings me straight to another, older, book by Erin Meyer, The Culture Map: Breaking Through the Invisible Boundaries of Global Business. It reads like a compressed tutorial of inter-cultural communication and decision making, although I have to admit it was almost more fun to naively learn all of this in the field than to have all the findings confirmed at a later point by the conclusions in this book.
I found it particularly interesting to see the historical context for some present cultural behaviour, by which I don’t mean the obvious teaching of Confucius for Chinese people but also current social traits in Europe dating back to the Roman Empire.
So when a Chinese colleague, who recently moved to Germany, described the German personality as a coconut after the first weeks of adjusting to life in Munich, it made me think that we should be providing this book as a compulsory read within the company, just to soften the blow.
Lastly, looking at how big data and analytics started to change our lives in many domains and found their way into sport in the classic Moneyball book, I believe that no other sport has been changed as drastically by a statistical approach to analytics as basketball.
Kirk Goldberry’s Sprawlball: A Visual Tour of the New Era of the NBA explains the dramatic change in the game by findings that, in hindsight, are so obvious that one can only wonder how we all didn’t see it coming in the 1990s when the GOAT Michael Jordan redefined the art of playing ball.
Goldberry explains the historical context for modern-day greats like LeBron James, James Harden and Steph Curry, while also giving a shout-out to my other personal favourite, Dirk Nowitzki, whose 2011 finals run will stay at the top of my sporting moments.
I just wish I could have told my 14-year-old self to stop practising baby hooks and post ups and go straight to 3-point drills.