Ciena recently revealed that operator, Trans Pacific Networks (RPN) deployed its coherent optical technology across the Echo and Tabua submarine cable infrastructure. Here, Thomas Soerensen, VP of Global Submarine Solutions at Ciena discusses the WaveLogic 6 Extreme technology, the shift to open cable systems, geopolitical influences on route planning, and the future of subsea capacity.
Can you tell us about the TPN partnership and what makes Echo and Tabua strategically important?
We've worked with Trans Pacific Networks (TPN) and Pioneer Consulting for quite some time now. We recently announced that TPN is deploying our optical solutions across the Asia-Pacific region, specifically powering TPN’s Echo and Tabua subsea cables, Pioneer Consulting is overseeing the route engineering, quality assurance, and project execution.
The industry is moving into this open cable architecture, whereby cable owners build an open system – they go to NEC, ASN or SubCom and build the pipes. Then as they get closer to the cables being ready for service, that's when they make a choice on the SLTE side.
What makes these two cables very special? Tabua is a cable system built by Google. It's the first ever continental US system that has two landing points in Australia. There are other cables on that route, but this one is by far the highest capacity cable with 16 fibre pairs. It stretches 13,300km, with landing points in Fiji and Hawaii.
The Echo cable is the first cable ever directly between Singapore and the US, which reaches 16,051km. That one fibre pair cable is built by NEC. TPN owns three fibre pairs on that cable. Reaching a distance of 16,000km in one hop makes it quite challenging, but our new technology is reaching those distances. We have done trials on that system already, and it's looking very promising.
Is AI reshaping subsea network architecture and design requirements?
I think the jury is still out when it comes to subsea cables as to whether or not this is AI-driven. AI is a relatively new thing, particularly in terms of subsea systems. Cables like Echo and Tabua were contracted years ago – Echo probably five years ago – so the conversation around AI wasn’t as prominent as it is now.
The developments we are seeing are certainly AI-driven now. What we are seeing in terms of what is being lit on these cables and technology being equipped to these fibres is definitely driven by AI. But the routing of these cables was decided a long time ago. That's probably much more driven by the bandwidth-intensive workloads that are associated with cloud services, which are still continuing to surge. That increases the need for high-capacity infrastructure.
What are the key technical advances in WaveLogic 6 Extreme?
Ciena was the first to introduce coherent technology into subsea 12 to 14 years ago. With that, we started this continued development of new technologies, making Ciena the first on 40G, on 100G, on 400G, on 800G. Now with WaveLogic 6 we're reaching speeds of 1.6 terabits on the shorter distances, and we like to say 800G everywhere, which means our latest technology can reach speeds of 800G anywhere, any distance that we know of today.
This is achieved mainly via our new three nanometer technology that we are the only ones in the industry using, and the DSP that we have developed in-house. That gives us speeds of up to 1.6 terabits per second. We have seen subsea systems in the North Sea where we have tested 1.6T. On Trans-Pacific distances, we sit somewhere between 800G and 1.1T. It also depends on how the cables actually perform.
The other thing made possible by the three nanometer DSP is transmission rates and combinations where we're going from 140 gigabaud to up to 200 gigabaud in increments of 0.1 dB. That combined with frequency between 400G and 1.6T gives us up to 26,000 different transmission modes we can utilise on every wavelength, maximising the spectrum and capacity on every cable.
One of the benefits compared to previous technology is the reduction in power consumption of 50% per bit. We also need much less space, and that again is facilitated by the three nanometer technology. Cable landing stations were built in the days where we had a cable coming in with four fibre pairs. Right now that same cable is 16 to 24 fibre pairs. If that gear had the same size as it did five years ago, we couldn't manage it in the cable landing stations.
How important is open architecture and spectrum sharing?
The open architecture started probably about 10 years ago. Till then, every time you built a cable system, it was turnkey – you went to a supplier and got SLTEs and the cable. About 10 years ago, we introduced this open cable concept where you don't buy the SLTEs as part of your supply contract with the wet plant suppliers. You just buy the wet plant.
One of the benefits is you don't want to make a technology choice five years before you need the equipment, because technology would have moved on. You're much better off waiting till a year, 18 months before your system is ready, until you make the choice. You go to the market – and obviously it's an open market – so you go to whoever you feel has the best product serving your requirements. 99 per cent of all cables being built today are open cable systems.
Spectrum sharing is a big thing. If you want to sell capacity on a network less than a full fibre pair, then you need spectrum sharing equipment to slice the fibre pairs up into smaller chunks. Spectrum sharing is definitely becoming pretty much a standard on every cable.
Are the specs changing with hyperscaler-led systems?
I wouldn't say it's changing specs. It's probably pushing specs. It's pushing the technology. They want the best cost per bit. How do you get the best cost per bit? Obviously more capacity. We are getting closer to the Shannon limit when we look at what we can actually do with each of the fibres.
A lot of the focus now is on the actual cables. Cables that were built were between four and six fibre pairs. Now with SDM technology, we're averaging around 18 fibre pairs per cable in what is actually getting deployed. The industry is incredibly active. We're averaging around 50 cables right now – that's systems over 1,000 kilometres. If you go five years back compared to 2026, we're doubling the number of cables but we are quadrupling the number of fibre pairs in each cable.
What's driving the focus on diverse routes like Echo and Tabua?
It is definitely more a resilience play today than a latency play. Most data centres, if you talk subsea intercontinental, latency is not the biggest factor. It is by far resilience.
Echo is the first cable directly between Singapore and the US. The shortest distance from Singapore to the US is not going east out of Singapore. The old traditional route would be going from Singapore towards Taiwan, through the Formosa Strait and then out to the US.
Right now, it is not very easy for non-Chinese companies to build cable systems through the South China Sea, and therefore the routing is going east of Borneo, east of Philippines, keeping on the Pacific side.
We're seeing Project Waterworth, the Meta system building around the world that avoids the Red Sea, avoids the South China Sea, avoids Indonesian waters. We have seen Google building Moya, which is the first South Africa to Australia cable system ever, which will allow Google to circumvent the globe without going through the Red Sea or Middle East region. Even if you look at the Atlantic, the traditional route 10 years ago was New York-London. Now the latest systems are much more south – into Portugal, into Spain. We are seeing that focus on resilience everywhere.
What are the main technical bottlenecks for even higher capacities?
Right now, it's the amount of power we can put into the cable. What the industry is looking at is C+L band technology. C+L band is where you utilise both the C band and the L band. C+L band technology has on the terrestrial side really increased in volume. That makes it more relevant again for subsea.
Another technology we're seeing is multi-core fibre. There's one system being deployed now, the CPU system Taiwan-Philippines-US, that utilises multi-core fibre. From our perspective, a multi-core fibre is just like getting more fibres, more glass, and we can connect to that. With the technology we can see in the horizon now, we can probably quite easily go from where we are now – 24 fibre pairs – to 24 fibre pairs with up to four cores, making effectively 96 fibres.
If we go a little further out, we're probably going to be seeing hollocore fibre. Hollocore is beginning its life on the terrestrial side. It is a very promising technology. I would be surprised if we do not see that in subsea. It's probably 10 years out. One of the issues with subsea is mechanical – a hollocore fibre has a very high risk of water ingress. Those are some of the issues we need to deal with.
What about automation and AI operations?
We have something called ADO, Automated Deployment Optimizer. With 26,000 different transmission modes, you can imagine the formula you would need to ensure you maximise each and every wavelength. We have developed a software tool as part of our Navigator NCS that can do all of that automatically. What used to take weeks in the industry now takes a couple of hours with a full channel plan, and it can even notice the traffic at the end. That's certainly one of the things hyperscalers are asking for. As we see more fibre pairs, more cables, speeds going up, we need more automation.
We're also working on AI Ops, which is an AI platform that helps with network management systems. We're beginning to get machine learning into understanding the networks better and helping us do fault-finding much quicker.
What's the next major inflection point for the subsea industry?
AI is an obvious inflection point and driver of changing customer needs and technological progression. The industry is working hard to keep capability ahead of demand and consider which routes will become more strategically important around the world.
Other inflection points are the permitting hurdles to consider and wider geopolitical context. Our industry has been impacted to some extent by geopolitics throughout the years, and the industry has been very good in adapting to the changes we have seen geopolitically. Where we are seeing more tangible change is permits.
The narrative of the transatlantic is that if you go 10 years back, permitting time was around 18 months. Now it's closer to 48 months to obtain permits for a transatlantic cable. We have various governments having increased focus on maritime domain awareness. You've seen the UK and Norway going together to monitor and survey critical infrastructure in the North Sea.
As a subsea industry, for many years we were yelling and screaming for more visibility into what we're doing. Now we have all of that, but it comes at a cost. Things take longer, there's more vetting, and many more parties are interested in what this industry is doing.
